This action might not be possible to undo. Are you sure you want to continue?

# THE EFFECTS OF METACOGNITIVE SCAFFOLDING AND COOPERATIVE LEARNING ON MATHEMATICS PERFORMANCE AND MATHEMATICAL REASONING AMONG FIFTH-GRADE

STUDENTS IN JORDAN

by Ibrahim Mohammad Ali Jbeili

September 2003

Thesis submitted in fulfillment of the requirements for the degree of Doctor of Philosophy

1

Table of Contents List of Tables List of Figures List of Appendices Acknowledgements Abstrak Abstract

vii ix

i

x xi xiii xvi

Chapter One

INTRODUCTION 1.1 Background to the Statement of the Problem 1.2 Statement of the Problem 1.3 Research Questions 1.4 Hypotheses 1.5 The Theoretical Framework 1.6 Significance of the Study 1.7 Operational Definitions LITERATURE REVIEW 2.1 Introduction 2.2 Objectivist Views Regarding the Learning/Teaching of Mathematics 2.2.1 Behaviorism and the Learning / Teaching of Mathematics 2.2.2 Gagne and the Learning / Teaching of Mathematics 2.2.3 Landa and the Learning / Teaching of Mathematics 2.2.4 Scandura and the Learning/Teaching of 2.3 Mathematics Constructivist Views Regarding the Learning/Teaching of Mathematics 2.3.1 Nature of the Learning Process and Construction of Knowledge 2.3.2 Piaget and the Learning / Teaching of Mathematics 2.3.3 Vygotsky and the Learning / Teaching of

Mathematics

1 1 9 15 16 17 19 21 24 24 25 25 27 29

Chapter Two

30 34 34 38 40 41

**2.3.4 Bruner and the Learning / Teaching of Mathematics
**

2

2.4

Metacognitive Strategies and the Construction of Knowledge 2.4.1 Regulation of Cognition 2.4.2 Metacognitive Strategies and Age 2.4.3 Metacognitive Scaffolding Cooperative Learning and Learning Mathematics with Understanding 2.5.1 Theoretical Perspective on Cooperative Learning 2.5.2 Elements of Cooperative Learning 2.5.3 Teacher’s Role in Cooperative Learning Cooperative Learning with Metacognitive Scaffolding and Learning Mathematics with Understanding 2.6.1 Conceptual Understanding 2.6.2 Procedural Fluency 2.6.3 Strategic Competence 2.6.4 Adaptive Reasoning 2.6.5 Productive Disposition Cooperative Learning with Metacognitive Scaffolding and Mathematical Reasoning Cooperative Learning with Metacognitive Scaffolding and Real-Life Problem Solving Cooperative Learning with Metacognitive Scaffolding and Motivation

43 45 47 48 52 53 56 57 60

2.5

2.6

2.7 2.8 2.9

63 65 67 70 72 74 77 80

Chapter Three

METHODOLOGY 3.1 3.2 3.3 3.4 3.5 Introduction Population and Sample Experimental Conditions Research Design Instructional Materials and Instruments 3.5.1 Instructional materials 3.5.1.1 Adding and Subtracting Fractions 3.5.1.2 Unit The Metacognitive Questions

83 83 83 84 86 88 88 88 89 90 90 92 96

**Cards 3.5.2 Instruments 3.5.2.1 The Mathematics Achievement 3.5.2.2 3.5.2.3
**

3

Test The Scoring of Mathematics Achievement Test The Metacognitive Knowledge

3.6

Questionnaire 3.5.3 Materials and Instruments Validity 3.5.4 Instruments Reliability Procedures 3.6.1 The Pilot Study 3.6.2 The Formal Study 3.6.3 Groups’ Equivalence 3.6.4 Teachers’ Training 3.6.5 Implementation of the Study

96 98 98 99 99 100 100 102

3.6.5.1

The Cooperative Learning with Metacognitive Scaffolding (CLMS) Method

102

3.6.5.2

The Cooperative Learning (CL)

106 107 108 110 110 110 111 112 113 116 116 117 117 121 121 126 127 133 133 139 139

3.7

Method 3.6.5.3 The Traditional (T) Method 3.6.5.4 Implementation Fidelity Data Analysis Procedure and Method 3.7.1 The pre-Experimental Study Findings 3.7.2 3.7.3 3.7.4 3.7.5 Analysis The Experimental Study Findings Analysis Justifications for using two-way MANCOVA / MANOVA Pearson’s Correlation Assumptions for MANOVA / MANCOVA

Chapter Four

RESULTS 4.1 4.2 4.3 Introduction The pre-Experimental Study Results 4.2.1 Statistical Data Analysis The Experimental Study Results 4.31 Testing of Hypothesis 1 4.3.2 Summary of Testing Hypothesis 1 (CLMS > CL

> T)

**4.3.3 Testing of Hypotheses 2 4.3.4 Summary of Testing Hypothesis 2 (CLMSH
**

>CLH >TH)

**4.3.5 Testing of Hypotheses 3 4.3.6 Summary of Testing Hypothesis 3 (CLMSL >
**

CLL> TL)

4.3.7 Testing of Hypotheses 4

4

4.3.8 Summary of Testing Hypotheses 4 (There are interaction effects between the instructional methods and the ability levels)

145

4.3.9 Summary of Findings to Research Questions 146 1–4 Chapter Five DISCUSSION AND CONCLUSIONS 5.1 Introduction 5.2 Effects of the Instructional Methods on Mathematics Performance, Mathematical Reasoning, and Metacognitive Knowledge 5.2.1 Effects of the Instructional Methods on Mathematics Performance 5.2.2 Effects of the Instructional Methods on Mathematical Reasoning 5.2.3 Effects of the Instructional Methods on Metacognitive Knowledge 5.3 Effects of the Instructional Methods on Mathematics Performance, Mathematical Reasoning, and Metacognitive Knowledge Based on Ability Levels

5.3.1 5.3.2 Performance of High-Ability Students Taught Via CLMS

150 150 152

152 155 160 163

164 167 168 170 171 173 176 178 180 182

Performance of High-Ability Students Taught Via CL Performance of Low-Ability Students Taught Via CLMS Performance of Low-Ability Students Taught Via CL Performance of Low-Ability Students Taught Via T

5.3.3

5.3.4

5.3.5

5.4 5.5 5.6 5.7 5.8

Interaction Effects Summary and Conclusions Implications for Educators Implications for Future Research Limitations of the Study

5

References 183 Appendices 199

6

List of Tables Table 3.1 3.2 3.3 3.4 4.1 4.2 Mechanisms for the Three Groups Research Design Pearson’s correlation among the three dependent variables (MP, MR, and MK) Pearson’s correlation among the covariates (pre-MP and pre-MR) and the dependent variables (MP, MR, and MK) Means and standard deviations on each dependent variable (pre-MP and pre-MR), by the groups Summary of multivariate analysis of variance (MANOVA) pre-MP and pre-MR results and follow-up analysis of variance (ANOVA) results. Means, standard deviations, adjusted means and standard errors for each dependent variable by the instructional method Summary of multivariate analysis of covariance (MANCOVA) results by the instructional method and follow-up analysis of variance (ANOVA) results Summary of post hoc pairwise comparisons Means, standard deviations, adjusted means and standard errors for each dependent variable for high-ability students by the instructional method Summary of multivariate analysis of covariance (MANCOVA) results by the instructional method and follow-up analysis of covariance (ANCOVA) results of comparing high-ability students across the three groups. Summary of post hoc pairwise comparisons between high-ability students across the three groups Means, standard deviations, adjusted means and standard errors for each dependent variable for low-ability students by the instructional method Page 85 86 113 115 118 120

4.3 4.4

122 124

4.5 4.6

125 128

4.7

130

4.8 4.9

131 134

4.10

**Summary of multivariate analysis of covariance (MANCOVA)
**

7

136

results by the instructional method and follow-up analysis of variance (ANOVA) results of comparing low-ability students across the three groups. 4.11 4.12 Summary of post hoc pairwise comparisons between low-ability students across the three groups Means, standard deviations, adjusted means and standard errors for each dependent variable by the interaction between the instructional methods and the ability levels (high-ability and low-ability) Summary of multivariate analysis of covariance (MANCOVA) results by the interaction effect and follow-up analysis of covariance (ANCOVA) results across the three groups. 137 140

4.13

142

8

List of Figures Figure 4.1 4.2 4.3 Interaction effect between the instructional method and the students’ ability levels on MP Interaction effect between the instructional method and the students’ ability levels on MR Interaction effect between the instructional method and the students’ ability levels on MK Page 143 144 145

9

List of Appendices Appendix Appendix A Appendix B Appendix C Appendix D Appendix E Appendix F Metacognitive Questions Cards Mathematics Achievement Test

Distribution of Scores across the Test Items Scoring Rubric

Page 200 206 212 213 214 218

Metacognitive Knowledge Questionnaire and Scoring Key Permission of Conducting Research in the First Public Educational Directorate Schools under the Jordan Ministry of Education Permission of Conducting Research in Irbid Governorate Schools under the First Public Educational Directorate

Appendix G

220

10

Acknowledgements In the name of Allah, Most Gracious, Most Merciful I am grateful for all the bounties that Allah has showered on me which enabled me complete this doctoral thesis. I also thank Allah for providing me with a supportive family and supportive colleagues and friends during my graduate studies. I would like to express my appreciation to all the individuals without whom the completion of this thesis would not be possible. First of all, my heartfelt thanks go to my thesis major supervisor, Associate Professor Dr. Merza bin Abbas, for his warm personality, continual and unwavering encouragements, support, tutelage, patience, and perseverance in guiding me through the entire research and thesis-writing process. My deepest thanks also go to my co-supervisor Associate Professor Dr. Wan Mohd Fauzy for his invaluable assistance. I would also like to express my particular thanks to the faculty and administrative staff of the Center for Instructional Technology and Multimedia, University of Science Malaysia, who provided facilities, and advice and support. My thanks also go to the administrative staff of the Institute of Post-graduate Studies, IPS, USM, for their assistance and support. I would also like to gratefully acknowledge the principals, teachers, and students of the primary schools which served as research sites: AlMuthana bin Harethah School, Huthaifa bin Alyaman School, and Abd Arrahman Alhalholi School. My profound gratitude goes to the Director of the Educational Development and Research Department, Jordan Ministry of Education and the Director of First Public Educational Directorate in Irbid Governorate for their assistance.

11

I am grateful to my colleagues, Dr. Sayed Anwar, Dr. Sharifa, Zainal, Oi, Husaini, and Aree in Center for Instructional Technology and Multimedia for their friendship over the past few years and the immense help during my research process. My gratitude also goes to the administrative staff of the University of Science Malaysia’s Library for their patience and assistance. My acknowledgement is also extended to Dr. Jeremy Kilpatrick, University of Georgia, Dr. Marjorie Montague, University of Miami, Dr. David Johnson and Dr. Roger Johnson, University of Minnesota, and Dr. Khattab Abu Libdeh, Jordan National Center for Educational Research and Development who provided valuable documents, articles, and suggestions throughout my thesis-writing. Last but not least, my affectionate thanks go to my family for their unfailing love, continual understanding, sacrifice, prayers and confidence, and selfless support: My parents, my brothers and sisters, and my wife and my mother in-law. I would like to express my gratitude to my father and my brothers who financially supported me during my graduate studies and patiently waited for me to finish my study. My mother, may Allah reward you for your patience and prayers before and during my graduate studies. I thank Allah for having a very understanding and loving wife. She has given me tremendous support and continuous encouragement during my thesis-writing. Words are inadequate to express my gratitude for their sacrifice, support, and patience and I love them with all my heart. “May Allah reward and bless all of my family members”. Ibrahim Mohammad Ali Jbeili

12

ABSTRAK Kesan Perancahan Metakognisi dan Pembelajaran Kooperatif Terhadap Prestasi Matematik dan Taakulan Matematik Di Kalangan Pelajar Tahun Lima di Jordan Tujuan penyelidikan ini ialah mengkaji kesan pembelajaran kooperatif berserta perancahan metakognisi terhadap (a) prestasi matematik (MP), (b) taakulan matematik (MR), dan (c) pengetahuan metakognisi (MK) di kalangan pelajar tahun lima di Jordan. menyelidiki Penyelidikan ini turut mengkaji kesan pembelajaran kooperatif berserta perancahan metakognisi terhadap skor-skor MP, MR, dan MK di kalangan pelajar berpencapaian tinggi dan rendah. Skor-skor MP, MR, and MK diukur melalui ujian pencapaian matematik dan soalselidik pengetahuan metakognisi. Reka bentuk eksperimen kuasi yang menggunakan reka bentuk factorial 3 x 2 telah digunakan dalam kajian ini. Faktor pertama ialah tiga paras kaedah pengajaran, iaitu (a) pembelajaran kooperatif berserta perancahan metakognisi (CLMS), (b) pembelajaran kooperatif tanpa perancahan metakognisi (CL), dan (c) pengajaran tradisional (T), iaitu kaedah pengajaran tanpa pembelajaran kooperatif atau perancahan metakognisi. Faktor kedua ialah pencapaian pelajar, iaitu Pencapaian Tinggi dan Pencapaian Rendah. Pembolehubah bersandar ialah skor-skor di dalam MP, MR, dan MK. Tiga sekolah rendah lelaki telah dipilih secara rawak dari sekumpulan empat puluh empat sekolah rendah yang mengajar matematik di dalam kelas-kelas heterogenus di mana pelajar tidak dikumpulkan atau ditindik mengikut keupayaan. 240 pelajar lelaki di dalam kelas tahun lima dari tiga buah sekolah rendah telah dipilih secara rawak, iaitu dua kelas dari setiap sekolah.

13

Satu ujian pra matematik telah ditadbirkan dan dua bulan sebelum kajian dimulakan kelas-kelas yang terlibat telah diperkenalkan dengan kaedah-kaedah CLMS and CL dan melaksanakan unit-unit latihan yang disediakan. Tumpuan kajian ini ialah pada unit “Penambahan dan Penolakan Pecahan” yang diajar di semua kelas selama 14 sesi pada penghujung semester pertama pada tahun akademik 2002 / 2003. Setiap kumpulan kooperatif CLMS terdiri dari dua pelajar pencapaian tinggi dan dua pelajar pencapaian rendah dan setelah mendengar pengenalan dari guru belajar secara kooperatif dan menggunakan kad-kad soalan metakognisi untuk memandu kerja-kerja serta latihan-latihan menyelesaikan masalah matematik yang disediakan. Dalam kaedah kooperatif ini, pembelajaran pelajar dibantu oleh perancahan oleh guru, oleh kad-kad soalan metakognisi dan oleh interaksi sesama pelajar. Ahli-ahli kumpulan CL juga terdiri dari dua pelajar pencapaian tinggi dan dua pencapaian rendah dan belajar secara kooperatif setelah mendengar pengenalan dari guru. Pelajar di dalam kumpulan kaedah T diajar secara lazimnya dan menyelesaikan masalah secara individu. Dapatan kajian ini menunjukkan bahawa pelajar kumpulan CLMS menunjukkan prestasi lebih tinggi yang berbeza secara signifikan dari kumpulan CL yang seterusnya menunjukkan prestasi lebih tinggi yang berbeza secara signifikan dari kumpulan T di dalam semua skor, iaitu MP, MR dan MK . Juga pelajar pencapaian tinggi di dalam kumpulan CLMS menunjukkan prestasi lebih tinggi yang berbeza secara signifikan dari kumpulan T dalam skor-skor MP, MR dan MK, serta prestasi lebih tinggi yang berbeza secara signifikan dari kumpulan CL dalam skor-skor MR dan MK. Pelajar pencapaian tinggi di dalam kumpulan CL pula menunjukkan prestasi lebih tinggi yang berbeza secara signifikan berbanding pelajar pencapaian tinggi kumpulan T dalam skor-skor MP, MR dan MK. Dapatan kajian juga menunjukkan bahawa pelajar pencapaian rendah di dalam kumpulan CLMS menunjukkan prestasi lebih tinggi yang

14

berbeza secara signifikan berbanding pelajar pencapaian rendah di dalam kumpulan CL dan kumpulan T dalam skor-skor MP, MR, dan MK. Pelajar pencapaian rendah di dalam kumpulan CL juga menunjukkan prestasi lebih tinggi yang berbeza secara signifikan berbanding pelajar pencapaian rendah kumpulan T dalam skor-skor MR dan MK. Dapatan kajian juga menunjukkan kesan-kesan interaksi yang signifikan dalam kumpulan CLMS di antara pencapaian pelajar dan skor-skor dalam MR dan MK dengan pelajar pencapaian rendah mendapatkan manfaat yang lebih dari kaedah yang digunakan.

ABSTRACT

15

The Effects of Metacognitive Scaffolding and Cooperative Learning on Mathematics Performance and Mathematical Reasoning among Fifth-Grade Students in Jordan The purpose of this study was to investigate the effects of cooperative learning with metacognitive scaffolding on (a) mathematics performance (MP), (b) mathematical reasoning (MR), and (c) metacognitive knowledge (MK) among fifth-grade students in Jordan. The study further investigated the effects of cooperative learning with metacognitive scaffolding and cooperative learning on high-ability and low-ability students’ achievement in MP, MR, and MK. The MP, MR, and MK scores were measured through a mathematics achievement test and a metacognitive knowledge questionnaire. A quasi-experimental study design that employed a 3 x 2 Factorial Design was applied in the study. The first factor was three levels of instructional method, namely, (a) cooperative learning with metacognitive scaffolding (CLMS), (b) cooperative learning without any metacognitive scaffolding (CL), and (c) traditional instruction (T) with neither cooperative learning nor metacognitive scaffolding. The second factor student ability levels, namely, high-ability and low-ability. The dependent variables were student achievement in MP, MR, and MK. Three male primary schools were randomly selected from forty four primary schools where mathematics was taught in heterogeneous classrooms with no grouping or ability tracking. 240 male students who studied in six fifth-grade classrooms were randomly selected from the three primary schools i.e., two classes from each school. A pre-mathematics achievement test was administered first, and then the CLMS and CL methods were introduced to the students with practice units two months before conducting the study. For the study, the focus was on the “Adding and Subtracting

16

Fractions” unit that was taught in all classrooms for 14 sessions at the end of the first semester for the academic year 2002 / 2003. In the CLMS method, after listening to their teacher’s introduction, students in small groups of two high-ability and two-lowability students worked cooperatively and used metacognitive questions cards to execute their mathematics exercises and solve mathematics problems. In this method, students’ learning was scaffolded by the teacher, the metacognitive questions, and the students’ cooperation. In the CL method, after listening to their teacher’s explanation, students worked cooperatively in small groups of two high-ability and two low-ability students. In the T method, students were taught in the usual manner and solved the mathematics problems individually. The results showed that overall the students in the CLMS group significantly outperformed the students in the CL group who, in turn, significantly outperformed the students in the T group in all measures. Additionally, the high-ability students in the CLMS group significantly outperformed their counterparts in the T group in MP, MR and MK, and significantly outperformed their counterparts in the CL group in MR and MK but not in MP. The high-ability students in the CL group in turn significantly outperformed their counterparts in the T group in MP, MR and MK. Also, the results showed that the low-ability students in the CLMS group significantly outperformed their counterparts in the CL group and in the T group in MP, MR, and MK. The lowability students in the CL group in turn significantly outperformed their counterparts in the T group in MR and MK but not in MP. Finally, the results showed significant interaction effects between student ability and the instructional method for the MR and MK scores with the low-ability students in group CLMS benefiting more than the high-ability students.

17

CHAPTER ONE

INTRODUCTION This study investigated the teaching of mathematics based on constructivist principles. The study focused primarily upon the investigation of the effects of cooperative learning with metacognitive scaffolding and cooperative learning on Jordan fifthgrade students’ mathematics performance, mathematical reasoning, and metacognitive knowledge in learning and solving problems involving the addition and subtraction of fractions. This first chapter of the study presents the background to the statement of the problem, specifies the statement of the problem and the purpose of the study, and describes its questions and hypotheses, and presents the study theoretical framework and the significance of the study. Finally, the chapter presents the operational definitions. 1.1 Background to the Statement of the Problem Children today are growing up in a world permeated by mathematics. The technologies used in homes, schools, and the workplaces are all built on mathematical knowledge. Many educational opportunities and good jobs require high levels of mathematical expertise. Mathematical topics arise in newspaper and magazine articles, popular entertainment, and everyday conversation. Mathematics is a universal, utilitarian subject, that is, so much a part of modern life that anyone who wishes to be a fully participating member of society must know basic mathematics. Mathematics also has a more specialized, esoteric, and esthetic side. It epitomizes the beauty and power of deductive reasoning. Mathematics embodies the efforts accumulated over thousands of years by every civilization to comprehend nature and

18

bring order to human affairs. For students to participate fully in society, they must learn mathematics with understanding, how to connect mathematical ideas, and how to reason mathematically. Students who cannot reason mathematically are cut off from whole realms of human endeavor. Students without mathematical understanding are deprived not only from opportunity but also from competence in everyday tasks (Kilpatrick et al., 2001). So mathematics instruction should emphasize such variables that result learning with understanding in order to meet the changing demands of the society. Mathematics instruction has moved through a series of development phases. The move from behaviorism through cognitivism to constructivism represents shifts in emphasis away from an external view to an internal view of learning. To the behaviorist, the internal processing is of no interest; to the cognitivist, the internal processing is only of importance to the extent to which it explains how external reality is understood. In contrast, the constructivist views the student as a builder of his knowledge (Jonassen, 1991). This turning point of learning processes asks for instruction that deals with students as builders not receivers of knowledge, students who construct knowledge through interaction and connecting their experiences with the current situations, and students who have learning strategies to help in building their knowledge and understanding. Thus, successful and effective mathematics instruction emphasizes the teaching of strategies that enable students to plan, monitor, evaluate, and then construct their own knowledge and understanding. Particularly, for Jordan educational system, Jordanian human resource based economy was hard hit in the wake of the 80s’ slump in the regional oil economy, which had during its boom given tangible spillover benefits to the country in the form of

19

remittances from Jordanian skilled workforce working in Gulf Cooperation Council Countries. The slump also caused the general education system and particularly mathematics education to gradually lose its utility. The technological revolution and growing use of modern technologies in the industries as well as in other employment sectors had changed the mathematical knowledge and skills requirements of labor markets (Ahlawat and Al-Dajeh, 1996). It became necessary for Jordan to upgrade the quality of school graduates in order to meet the changing demands of the domestic labor market and to maintain its skilled workforce advantage in the region wide labor market. Under these circumstances, Jordan in 1989 launched a comprehensive 10-year-long Education Reform Plan (ERP) to overhaul the general education system. Mathematics education was one of the core subjects that received a lot of attention. The overarching objective of the reform plan was to enhance student achievement levels. The key reform elements were reconstructing the curricula, designing new textbooks and instructional materials, and conducting in-service teacher training in classroom applications of innovative instructional methods for using new textbooks and materials (Ahlawat and Al-Dajeh, 1996). To determine if the mathematics education reform has had the desired effects, in 1995, the National Centre of Human Resources Development (NCHRD) conducted a study to investigate the changes in mathematics achievement levels after five years of reform (Ahlawat and Al-Dajeh, 1996). The findings showed that there was a significant improvement in the whole field of mathematics achievement. The improvement, however, related to the routine mathematics concepts, procedures, and problem solving. The newly designed textbooks and in-service training did not cover

20

high-level cognitive skills, analytical thinking, and reasoning that help students to build their knowledge and develop understanding (Innabi, Hanan; Kaisee, and Hind, 1995). Ahlawat and AL-Dajeh (1996) indicate that the mathematics materials after reform covered only three cognitive skills (conceptual knowledge, procedural knowledge, and problem solving), and the improvement of conceptual understanding was the weakest according to the post-reform achievement tests. Moreover, from the analysis of individual items of students’ responses, there was significant deterioration in performance particularly with topics that involve abstract theoretical concepts (Ahlawat and Al-Dajeh, 1996). This indicates that mathematics teachers and materials developers in Jordan concentrate on learning procedures exclusively and do not pay attention to the teaching of strategies that help students to build and develop conceptual understanding and reasoning. In the last year of the Jordan Education Reform Plan, the Third International Mathematics and Science Study–Repeat (1999) was conducted to compare mathematics and science eighth-grade students’ achievements among 38 nations. Jordan was among the 38 nations that participated in the study. TIMSS-R assessed five mathematics content areas: fractions and number sense, measurement, data representation (analysis, and probability), geometry, and algebra. The findings showed that the average mathematics achievement across the all five mathematics areas of Jordanian students was 428, which was lower than the international average which was 487. The lowest average was in fractions and number sense with an average score of 432, while the international average was 487. In terms of ranking, Jordan was placed at number 32 out of 38. Singapore was ranked first with an average of 604 points. In terms of quality of the scores (students scored 616 or higher), 46 percent of

21

the scores of the Singapore students were in the top ten percent compared to only 3 percent for Jordan. For Jordanian eighth-grade students, for example, on an item testing for concept

knowledge of fractions that asked students to shade

3 of 24 cells correctly, only 8

30.9% students responded correctly, while the rest gave different incorrect responses like for example, shading 3, 8, 11 cells, or giving unclear responses. Abu Libdeh (2000) indicates that mastering this item falls in the third-grade with 50% accuracy as criterion and increases to 80% in the fifth-grade. So this finding shows that 30.9% of eighth-graders accuracy level to this item is below the desired level. Many factors may affect and contribute to the students’ mathematical understanding and achievement. The understanding of mathematics as an academic subject and its perceived importance in school and life plays a very crucial role. For example, Singapore values mathematics very highly and its primary schools systematically teach aspects of mathematics normally reserved for middle schools or junior high schools (Kaur and Pereira, 2000). Studies by Sternberg and Rifkin (1979) and Thornton and Toohey (1985) have indicated that young children can benefit from using sophisticated strategies in solving mathematics problems. However, the current practice in most schools in Jordan has been to underestimate students’ real abilities to learn mathematics. The nature of mathematics is also called into question. It is usually classified as one of the usual science subjects together with physics and chemistry but it is not taught as a science subjects. It is conceived as consisting of numbers, rules, formulas, and algorithms, and the teaching has focused on the acquisition of procedures (Gagne,

22

1985) and algorithms and heuristics (Landa 1983) in solving routine and novel problems. Also Romberg (1988) indicates that mathematics in many cases is divorced from science and other disciplines and then separated into subjects such as arithmetic, algebra, geometry, trigonometry, and so on. Within each subject, ideas are selected, separated, and reformulated into a rational order. This is followed by subdividing each subject into topics, each topic into studies, each study into lessons, and each lesson into specific facts and skills. This fragmentation of mathematics has divorced the subject from reality and from learning with understanding. Such essential characteristics of mathematics as abstracting, inventing, reasoning, and applying are also often lost from Jordanian textbooks and teaching methods (Abu Libdeh, 2000). Thus, the learning of mathematics becomes the learning of isolated facts and skills that according to Gipps and McGilchrist (1999) “quickly disappear from the memory because they have no meaning and do not fit into the student’s conceptual map. Knowledge learned in this way is of limited use because it is difficult for it to be applied, generalized or retrieved” (p 47). By scrutinizing Jordanian fifth-grade mathematics textbook and teacher’s guide, it is clear that the Jordanian mathematics classrooms are dominated by seatwork, homework, and review. For example, instructions of teaching fractions emphasize the demonstration of mathematical facts, computation skills, and procedures used to solve routine problems. These instructions lead to teaching strategies that require mathematics teachers to concentrate on mastering procedures needed to solve routine tasks and problems. In addition, these instructions necessitate mathematics teachers to fragment mathematics materials and to include many topics, with a considerable amount of repetition of content which is divorced from reality. Therefore, mathematics instruction in Jordan often concentrates on teaching mathematical facts, skills, and

23

procedures with much less concentration on developing analytical thinking like, for instance, reasoning and metacognitive strategies. This kind of teaching promotes a sequential mastery of knowledge, with the teacher as a giver and the student as a receiver, but does not promote the view that students have potential abilities to build, plan, monitor, reason, and evaluate their knowledge (Wilkins and Jesse, 2000). Also this teaching does not promote values and other knowledge associated with mathematical proficiency as to be discussed in this study. The traditional sequence of teaching placing value, digit numbers, fractions, ratio decimals, percentages, and geometry tends to fractionalize mathematical knowledge instead of integrating and connecting it. This instruction influences students’ views of mathematics and may make them unable to transfer what they have learned to new situations in the real life. Moreover, this instruction does not encourage students to invest their reasoning, connection, and metacognitive strategies in their learning. That is, students often receive problem-solving procedures from the teacher without actual participation in planning, monitoring, and evaluating their learning. In this environment there is no opportunity for students to connect mathematical relationships between what they have learned and the current situations in which these connections enable them to recognize the importance of mathematics in all parts of life (Baroody, 1998), and learn mathematics with understanding (Carpenter and Lehere, 1999; Kilpatrick et al., 2001). Mathematics teachers in Jordan primary schools generally teach their students by means of conventional instructional methods. They select a set of mathematical problems, demonstrate the necessary steps leading to their current solutions and their students then follow the same steps in finding solutions to similar problems. This

24

pedagogical approach may be effective for high-ability students, but it may not work with low-ability students. The explanations for low mathematics achievement of lowability students could be that they are not taught the appropriate strategies, cannot selfregulate the study strategies, and do not understand how to apply these strategies (Simpson, 1984). The TIMSS-R (1999) findings reveal that Jordanian students’ achievement is still very low even after ten years of educational system reform. Moreover, students’ understanding of mathematics has also not improved. As Abu Libdeh (2000) found, a great deal of errors students gave is related to undetermined errors (unspecific errors that comprise unrelated responses, deleted responses, or unreadable responses). This means that most Jordanian eighth-grade students are lacking in or do not have mathematical understanding, strategic competence, and adequate reasoning skills for mathematics; therefore they responded with unrelated, unreadable, or other meaningless responses. While students’ difficulties in doing mathematics are partly attributed to misconceptions or shallow conceptions of domain knowledge (Feltovich et al., 1996), they are, to a greater extent, due to a lack of metacognitive strategies (Brown, 1987). Comparisons of good and poor comprehenders have consistently shown that poor comprehenders are deficient in the use of metacognitive strategies (Golinkoff, 1976; Meichenbaum, 1976; Ryan, 1981). So Jordanian students need to be taught mathematics through an effective instruction that enables them to acquire and apply metacognitive strategies, reason mathematically, and thus learn mathematics with understanding. In other words, students have to be taught and supported to plan, formulate and represent the mathematical problems, analyze and identify the mathematical variables,

25

connect the relationships among the mathematical variables, ask themselves questions regarding mathematical situations, reason mathematically, evaluate their strategies and outcomes (Kilpatrick et al., 2001; King, 1992), and to work cooperatively to learn with understanding (Palincsar and Brown, 1984). Particularly, students need to learn how to learn, that is, to be metacognitively trained. To date, insufficient attention has been given to the important role the metacognitive strategies play in improving mathematics performance and mathematical reasoning. National Research Council (NRC, 1998) declares that effective mathematics instruction must start from the earliest grades. Simply offering or even requiring all students to take a standard first year course in eighth-grade is no assurance that they will succeed. This method virtually guarantees failure for a large number of students. Instead, as Kilpatrick et al. (2001) suggest that students must begin to acquire the rudiments of learning with understanding in the earliest grades, as part of a comprehensive method to developing their mathematical proficiency. Moreover, Abu Libdeh (2000) found that a great deal of Jordanian eighth-graders’ mathematical errors and misunderstanding refer to the topics they learned in the fifth-grade. 1.2 Statement of the Problem New applications and new theories have given emphasis to instructional methods that play an important role in developing the learning of mathematics. Documents such as those produced by the National Council of Teachers of Mathematics (NCTM, 1989, 1991, and 1995) and the National Research Council (NRC, 1989) suggest that traditional mathematics instruction has been challenged by the changing expectations of the skills and knowledge of workers, and therefore, mathematics instruction should shift from concentrating on the products to the learning processes that comprise

26

learning strategies, planning, monitoring, evaluation, and reasoning. In other words effective mathematics instruction gives special attention to teach students how to learn and how to reason and evaluate their learning and solution processes. This view has sparked debate in the mathematics education community around the nature of the effective mathematics instruction and the experiences students need to learn mathematics with understanding. Some advocate a focus on conceptual, procedural, and reasoning competences (Mathematics framework for California Public Schools, 1999), while others argue for a focus on applications of mathematics and the study of realistic mathematics (Apple, 1992). Others state that learning mathematics with understanding brings together problem solving, reasoning, and criticalness. Frankenstein (1990), for example, calls for a mathematics method that emphasizes teaching mathematics through its applications with a goal of helping students become critical of the uses of mathematics in society. Schoenfeld (1985) argues that effective mathematics instruction must require students to understand mathematical concepts and methods, recognize relationships and think logically, and apply the appropriate mathematical concepts, methods, and relations to solve problems. Still others argue that learning mathematics with understanding (Kilpatrick et al., 2001) requires mastering and transferring the mathematical proficiency strands which are: conceptual understanding, procedural fluency, strategic competence, adaptive reasoning, and productive disposition in an integrated manner. Others (Mugney and Doise, 1978; Rogoff, 1990; Vygotsky, 1978) concentrate on cooperative learning to learn mathematics with understanding. Finally, Flavell et al. (1970) and Brown (1987) focus on metacognitive strategies to be taught to enable students to learn mathematics with understanding.

27

There is general agreement that learning mathematics with understanding involves more than competency in basic skills. Much more than mastering arithmetic and geometry, learning mathematics with understanding deals with conceptual understanding, procedural fluency, and reasoning (Kilpatric et al., 2001). Learning mathematics with understanding is more than learning the rules and operations that students learn in school. It is about connections, seeing relationships, and knowledge reconstruction in everything that students do (Brown et al., 1994). In summary, learning mathematics with understanding is about acquiring and improving conceptual understanding and procedural fluency (mathematics performance), mathematical reasoning, and activation of metacognitive knowledge. One way of supporting and improving students’ mathematics performance, mathematical reasoning, and metacognitive knowledge is through the provision of metacognitive strategies, which is an instructional method that concentrates on monitoring one’s current level of understanding and decides when it is not adequate (Bransford et al., 2000). It helps students to manage their thinking, recognize when they do not understand something, and adjust their thinking accordingly (Schoenfeld, 1992). In other words, metacognitive strategies guide students to think before, during, and after a problem solution. It begins by guiding students to plan for selecting the appropriate strategy to accomplish the task, and then continues as they select the most effective strategy and afterward evaluate their learning process and outcomes (Hacker, 1998). Metacognitive strategies according to Piaget’s cognitive development stages (1970) require abstract thinking that students become proficient in when they reach the formal operation stage (12 years and above). Young students, for example, 11 year

28

olds need to be supported, guided, or pushed to be metacognitive thinkers. Vygotsky (1978) explains the differences between students’ current abilities and their potential development as the distance between the actual students’ independent level and their potential level under guidance, support, or in collaboration with more capable peers. Scaffolding provides an opportunity for students to develop knowledge and skills beyond their independent current level, and this closes the distance between what is and what is possible. That is, with scaffolding, students are supported to go beyond their current thinking, so that they continually increase their capacities (Schofield, 1992). Researchers (e.g., Palincsar and Brown, 1984; Wood et al., 1976) have investigated the role of scaffolding to facilitate student comprehension, understanding, and reflection on complex tasks. In the studies of Palincsar and Brown (1984), Palincsar (1986), and Palincsar et al. (1987), scaffolding involved modeling and dialogue to enhance comprehension monitoring and strategy use. Scardamalia et al. (1984) provided coaching through question prompts, while King (1991a; 1992) modeled and guided students to use self-generate questions. These scaffolding strategies were shown to improve students’ cognition by activating their learning, enhancing knowledge retrieval, comprehension, and metacognition by making their thinking explicit and guiding them to monitor their understanding. Among the strategies of improving students’ mathematics performance, mathematical reasoning, and metacognitive knowledge is a recommendation for using cooperative learning (National Council of Teachers of Mathematics [NCTM], 1989; Kramarski, 2001). According to Vygotsky (1978) learning with understanding occurs within a social context. When students interact with each other, they typically will learn,

29

receive feedback, and be informed of something that contradicts with their beliefs or current understanding. This conflict often causes students to recognize and reconstruct their existing knowledge (Rogoff, 1990). Cooperative learning has been strongly recommended to be used in improving students’ cognitive performance, social relationships, and metacognitive knowledge (Dansereau, 1988; Paris and Winograd, 1990; Weinstein et al., 1994). The report of the National Governors’ Association (Brown and Goren, 1993) indicated that within cooperative learning setting, mixed ability students work together to solve problems and complete tasks. In this setting, low-ability students have the opportunity to model the study skills and work habits of more proficient students. In the process of explaining the material, high-ability students often develop greater mastery themselves by developing a deeper understanding of the task. However, there still exists uncertainty as to the mechanism by which improving students’ mathematics performance, mathematical reasoning, and metacognitive knowledge occurs within various cooperative learning environments. Does cooperative learning alone improve students’ mathematics performance, mathematical reasoning, and metacognitive knowledge? Or their cooperation needs to be structured and guided? If metacognitive strategies are provided to guide students’ cooperation, are students able to apply metacognitive strategies on their own, or do they need external scaffolding to do so? Do high-ability students benefit more than low-ability students from metacognitive scaffolding method? Elawar (1992) observed that lowability students are often found to be confused when they confront a mathematical problem and they are unable to explain the strategies they employ to find a correct solution. Costa (1985), Sternberg (1986 b) and Elawar (1992) indicated that lowability students generally lack well-developed metacognitive skills.

30

Although numerous research studies have been conducted on the separate effects of metacognitive strategies or cooperative learning on mathematics achievement, attitudes, and self-efficacy, no study was found that addresses the effects of cooperative learning with metacognitive scaffolding and cooperative learning on highability and low-ability students’ mathematics performance, mathematical reasoning, and metacognitive knowledge. Thus, the purpose of this study was to find out the extent to which the cooperative learning with metacognitive scaffolding and the cooperative learning methods could play an important role in improving Jordanian fifth-grade students’ mathematics performance, mathematical reasoning, and metacognitive knowledge. Particularly, the study was conducted to investigate if there were any significant differences in mathematics performance, mathematical reasoning, and metacognitive knowledge levels between students taught via the cooperative learning with metacognitive scaffolding instructional method (CLMS), students taught via the cooperative learning alone instructional method (CL), and students taught via the traditional instructional method (T). The study also examined the effects of the instructional methods on highability and low-ability students’ mathematics performance, mathematical reasoning, and metacognitive knowledge. As such, the study was focused on the following questions:

31

1.3 Research Questions 1. Would students taught via CLMS instructional method perform higher than students taught via CL instructional method who, in turn, would perform higher than students taught via T instructional method in (a) mathematics performance (MP), (b) mathematical reasoning (MR) and (c) metacognitive knowledge (MK)? 2. Would high-ability students taught via CLMS instructional method perform higher than high-ability students taught via CL instructional method who, in turn, would perform higher than high-ability students taught via T instructional method in (a) mathematics performance (MP), (b) mathematical reasoning (MR) and (c) metacognitive knowledge (MK)? 3. Would low-ability students taught via CLMS instructional method perform higher than low-ability students taught via CL instructional method who, in turn, would perform higher than low-ability students taught via T instructional method in (a) mathematics performance (MP), (b) mathematical reasoning (MR) and (c) metacognitive knowledge (MK)? 4. Are there interaction effects between the instructional methods and the ability levels (high-ability and low-ability) in mathematics performance, mathematical reasoning, and metacognitive knowledge?

32

1.4 Hypotheses Based on the research questions the following hypotheses were formulated: 1. Students taught via CLMS instructional method will perform higher than students taught via CL instructional method who, in turn, will perform higher than students taught via T instructional method in (a) mathematics performance (MP), (b) mathematical reasoning (MR) and (c) metacognitive knowledge (MK). 2. High-ability students taught via CLMS instructional method will perform higher than high-ability students taught via CL instructional method who, in turn, will perform higher than high-ability students taught via T instructional method in (a) mathematics performance (MP), (b) mathematical reasoning (MR) and (c) metacognitive knowledge (MK). 3. Low-ability students taught via CLMS instructional method will perform higher than Low-ability students taught via CL instructional method who, in turn, will perform higher than low-ability students taught via T instructional method in (a) mathematics performance (MP), (b) mathematical reasoning (MR) and (c) metacognitive knowledge (MK). 4. There are interaction effects between the instructional methods and the ability levels (high-ability and low-ability) on mathematics performance, mathematical reasoning, and metacognitive knowledge (MK).

33

1.5 The Theoretical Framework The theoretical base for this study comes from Piaget (1970) and Vygotsky (1978). According to the constructivist paradigm, students learn because they have taken prior knowledge and have reworked the new information into their current schema. A schema consists of the pieces of knowledge already present in the person. The processes, in Piagetian terms, that rework new information and incorporate it to prior knowledge are called assimilation and accommodation. When a new experience is incorporated into prior knowledge it is assimilated. Accommodation occurs when the new knowledge alters the knowledge, or schema (DeLay, 1996). Piaget (1970) believes that individuals work with independence and equality on each other’s ideas, so when the student is opposed new knowledge and interacts with others he or she encounters something that contradicts his or her believes or current understanding. This is what Piaget calls “cognitive conflicts” (Mugny and Doise, 1978). This conflict results a case of disequilibrium. Working cooperatively and activating metacognitive strategies such as planning, monitoring, and evaluation are likely to enhance students to assimilate or accommodate their knowledge and therefore reequilibrate their thinking. When students employ their metacognitive strategies, they are more than likely enhanced to revise, evaluate, and guide their ways of thinking to provide a better fit with reality.

34

While there is a general consensus that metacognitive strategies are developed with age, the developing mind does not develop in isolation, but within a social, cultural and linguistic environment such as in the case of the interaction with peers and adults. The need to explain and justify to others, makes reflection on ones own thought thus, developing metacognitive strategies (Pellegrini et al., 1996). Becoming more reflective and metacognitive enables students to provide for themselves the supportive and scaffolding role originally assigned to the adult or peer (Brown, 1987). Vygotsky (1978) believes that there is a hypothetical region where learning and development best take place. He identifies this region as the zone of proximal development. This zone is defined as the distance between what an individual can accomplish during independent problem solving, versus what can be accomplished with the help of an adult or a more capable member of a group. This is often a higherability individual. With cooperation, direction, or help, the individual is better able to solve more difficult tasks than he or she could independently. Furthermore, Vygotsky (1978) suggests that an active student and an active social environment cooperate to produce developmental change. The student actively explores and tries alternatives with the assistance of a more skilled partner, as in an instructor, or a more capable peer. The teacher and the partner guide and structure the students’ activity, scaffolding their efforts to increase current skills and knowledge to a higher competency level. Scaffolding is the support during a teaching session, where a more skilled partner (adult or peer) adjusts the level of assistance given based on the level of performance indicated by the student. A greater level of support is offered if the task is new, and less is provided as competency grows (Berk and Winsler 1995). The student is able to move forward and continues to develop new capabilities.

35

Therefore, guidance i.e., cooperatively and metacognitively, should be provided to support both cognition and metacognition. Cognition refers to domain-specific knowledge and strategies for information and problem manipulation (Salomon et al., 1989 and Schraw, 1998), and metacognition includes knowledge of cognition and regulation of cognition (Jacobs and Paris, 1987), such as planning, monitoring, and evaluation. The two constructs are interrelated. Although metacognitive knowledge may be able to compensate for absence of relevant domain knowledge, its development may also depend on having some relevant knowledge of the domain (Garner and Alexander, 1989). Thus, this study investigated the effects of cooperative learning with metacognitive scaffolding and cooperative learning methods on mathematics performance, mathematical reasoning, and metacognitive knowledge. That is, on learning with understanding where students assimilate or accommodate their mathematical knowledge. 1.6 Significance of the Study The research on metacognition, scaffolding, and cooperative learning strategies to enhance mathematical reasoning and understanding is based on meaningful learning. The information age has challenged educators to reexamine the role of the student and of instruction from the constructivist perspective. As the student’s role changes from a passive knowledge recipient to an active meaning constructor, reasoning, planning, monitoring, and evaluation have a significant value in instruction and play a significant role in understanding particularly in subjects based on proof like mathematics. Since learning mathematics with understanding requires skills more than

36

conceptual understanding and procedural fluency (Kilpatrick et al., 2001), cooperative learning with metacognitive scaffolding focuses on helping students to ask metacognitive questions that guide them to plan, understand, monitor, and evaluate and reason their learning, not just guides them to master mathematical procedures. In other words cooperative learning with metacognitive scaffolding focuses on helping students to be metacognitively prepared to solve mathematical problems with understanding and to plan, monitor, evaluate, and reason their solutions. It is hoped that the findings of this study will contribute to further understanding of the role of cooperative learning with metacognitive scaffolding and cooperative learning strategies in improving mathematics performance, mathematical reasoning, and metacognitive knowledge. If the use of cooperative learning with metacognitive scaffolding and cooperative learning methods proves its effectiveness in improving mathematics performance, mathematical reasoning and metacognitive knowledge, teachers in Jordan will have additional instructional methods that can be used to support students’ learning with understanding. Moreover, this will help educators in Jordan in their search for an effective and efficient pedagogical strategy or model for improving learning with understanding.

37

1.7 Operational Definitions Metacognition: The processes of considering and regulating student’s own learning that, include planning, monitoring, and evaluation of the student’s current and previous knowledge. These processes are activated before, during, and after the problem solution. Scaffolding: A technique in which the teacher provides instructional support as students learn to do the task, and then gradually shifts responsibility to the students. In this manner, the teacher enables students to accomplish as much of a task as possible without peer assistance. Cooperative Learning with Metacognitive scaffolding Method (CLMS): An instructional method in which, the high-ability and low-ability students work together in groups of four members (two high-ability and two low-ability students) to solve a problem or complete an assignment. In this method, the teacher, the metacognitive questions card, and the students’ interaction provide metacognitive scaffolding to students in the form of planning, monitoring, end evaluation in performing a given task. The teacher’s metacognitive scaffolding is gradually reduced as the students are able to accomplish the task.

38

Cooperative learning Method (CL): An instructional method in which, high-ability and low-ability students work together in groups of four members (two high-ability and two low-ability students) to solve a problem or complete an assignment. The teacher is allowed to assist the groups but the groups and the teacher are not provided with any metacognitive questions card. Traditional Instructional Method (T): An instructional method in which, the teacher explains and manipulates the mathematics concepts and procedures to the whole class. In this method, the teacher’s teaching time is about 35 minute out of 45 minute of the session’s time. The teacher’s concentration in this method is on mastering the task and developing specific skills. High-ability students: Students whose average scores on mathematics performance and mathematical reasoning measured by the pre-test are above the median. Low-ability students: Students whose average scores on mathematics performance and mathematical reasoning measured by the pre-test are below the median. Mathematical reasoning (MR): The student’s ability to make a decision about how to approach the mathematics problem, select or generate the appropriate tactic to solve the problem, support the solutions with evidence, and to generalize his solution processes to different situations. Conceptual understanding (CU):

39

The student’s ability to connect new mathematics ideas with ideas he has been known, represent the mathematical situation in different ways, and to determine similarities/differences between these representations. Procedural fluency (PF): The student’s ability to use mathematics procedures appropriately, flexibly, and accurately. Mathematics performance (MP): The students’ score in conceptual understanding and procedural fluency items on the mathematics achievement test. Mathematics achievement test: A mathematics test which assesses students’ conceptual understanding, procedural fluency, and mathematical reasoning simultaneously. Metacognitive Knowledge Questionnaire: A questionnaire consists of 15 items that assesses students’ planning, monitoring, and evaluation simultaneously , 5 items were designed to assess each skill. Primary government schools: Schools established by the Jordan Ministry of Education where students study from the first to the tenth-grade. These schools are not coeducational, so there are male schools and female schools.

40

CHAPTER TWO LITERATURE REVIEW 2.1 Introduction To understand how someone learns mathematics is an important matter. Understanding this serves educators to determine what and how they should teach. While understanding how someone learns mathematics is a difficult task, the study of psychology offers many contributions and deep understanding of how students learn mathematics. So reviewing the theories according to various psychological perspectives contributes to the understanding of how mathematics learning occurs on one hand, and serves in the understanding of how the teaching of mathematics should be conducted on the other hand. Mathematics is generally accepted as a very important school subject and thus the teaching and learning of mathematics have been intensively studied and researched over the past six decades. The study of the teaching of mathematics is always based on the conception of learning held by the researcher as well as the mental tasks believed to be necessary for performing mathematical tasks. This chapter reviews the

**paradigms, theories, and models of learning based on the literature currently available
**

41

to identify the theory, model, and variables most promising for use in improving the teaching of mathematics. The chapter also discusses the review of related literature on metacognitive scaffolding as well as cooperative learning. Then the chapter describes the mathematical proficiency model and discusses the role of metacognitive scaffolding and cooperative learning plays in improving the teaching of mathematics. The chapter continues with a discussion on mathematical reasoning and describes the role of cooperative learning and metacognitive scaffolding plays in improving students’ mathematics performance, mathematical reasoning, real-life problem solving, and motivation.

2.2 Objectivist Views Regarding the Learning / Teaching of Mathematics The objectivist theories postulate that knowledge exists independently of the student, and then becomes internalized as it is transferred from its external reality to an internal reality of the student that corresponds directly with outside phenomenon with the mind acting as a processor of input from reality (Driscoll, 1994). Meaning is derived from the structure of reality, with the mind processing symbolic representations of reality (Jonassen, 1991). This belief is very popular among educators and researchers and has produced numerous theories and models as represented by behaviourism, and cognitivism as represented by the theories of Gagne, Landa, and Scandura. 2.2.1 Behaviorism and the Learning / Teaching of Mathematics The learning paradigm of behaviorism represents the original Stimulus-Response (SR) framework of behavioral psychology: Learning is the result of associations forming between stimuli and responses. Such associations become strengthened or weakened

42

by the nature and frequency of the S-R pairings. The paradigm for S-R theory is trial and error learning in which certain responses come to dominate others due to rewards. The hallmark of behaviorism is that learning could be adequately explained without referring to any unobservable internal states. The behaviorists’ earlier studies concentrated on animals before becoming interested in human thinking. Thorndike (1932) states that in any given situation an animal has a number of possible responses, and the action that would be performed depends on the strength of the connection or bond between the situation and the specific action. The bonds that go together should be taught together. In pedagogical terms, this yields a drill and practice mode of instruction. At elementary school age for Thorndike, the rules of arithmetic are said to have not been known. The purpose of instruction in mathematics is thus seen to be one of drilling into the student the necessary rules and connections until sufficient responses are obtained. Thorndike explains this in his law of effect: “When a modifiable connection between a situation and a response is made and is accompanied or followed by a satisfying state of affairs, that connection’s strength is increased.” (p. 60). Skinner (1968) further argues that an organism learns mainly by producing changes in response to its environment. In other words, learning is characterized by changes in behavior. This may seem to be a simple truism except for the fact that Skinner argues that a change in behavior is the only characteristic of learning. He explicitly rejects such concepts as purposes, deliberations, plans, decisions, theories, tensions, and values as mentalism. Since these concepts are non-physical and therefore cannot be measured, weighed, and counted, he refers to them as pre-scientific and says learning need to move beyond such ideas and develop a true technology of behavior.

43

According to behaviorist principles, all learning processes are fully controlled by the teacher. So the teacher has to understand all of the students’ behaviors and subbehaviors involved in the task, as well as the characteristics of the students. Also the teacher has to create an instructional situation that requires students to practice the appropriate behaviors, in proper sequence and with appropriate reinforcement, gradually building more and more behaviors until the target behavior is achieved. This process requires a great deal of time for complex, intricate tasks such as data classification. The nature of mathematics as represented by behaviorism portrays mathematics not as a product of human creation but, instead, as existing external to the human mind. Tiene and Ingram (2001) assert that behaviorism is unable to effectively address the critical issue like how students think, understand, reason, and build knowledge. Students are more than just the sum total of the behaviors that they engage in. Students make plans, remember things, forget things, solve problems, hypothesize, and much more. These aspects of cognition could not be fully understood just by looking at behavior. Moreover, the role of the student in this environment is passive, namely, it is teacher-centered where the teacher selects, explains, demonstrates, and evaluates the instructional activities.

2.2.2 Gagne and the Learning / Teaching of Mathematics Gagne (1985) indicates that there are several different types of learning, and each type requires different types of instruction. He classifies human learning into five domains: intellectual skills, motor skills, cognitive strategies, verbal information, and attitudes. Different internal conditions such as acquisition and storage of prior capabilities, and external conditions such as the various ways that instructional events outside the student function to activate and support the internal processes of learning are

44

necessary for each type of learning. For example, for cognitive strategies to be learned, there must be an opportunity to practice developing new solutions to problems; to learn attitudes, the student must be exposed to a credible role model or persuasive arguments (Aronson et al., 1983). Gagne suggests that learning tasks for intellectual skills can be organized in a hierarchy according to complexity: discriminations, concept formation, rule application, and problem solving (facts, concepts, principles, and problem solving) (Aronson et al., 1983). In other words, the sequence in learning is from bottom up, that is, from simple to complex (Gagne, 1985). Gagne asserts that the significance of the hierarchy is to identify prerequisites that should be completed to facilitate learning at each level. Doing a task analysis of a learning task identifies the prerequisites. He determines two types of prerequisites: essential prerequisites which are the subordinate skills that must be previously learned to enable the student to reach the objective, and supporting prerequisites which are useful to facilitate learning but are not essential for the learning to occur. He adds that learning hierarchies provide a basis for the sequencing of instruction (Aronson et al., 1983). In addition, Gagne outlines nine instructional events that provide the external conditions of learning and corresponding cognitive processes: gaining attention (reception), informing students of the objective (expectancy), stimulating recall of prior learning (retrieval), presenting the stimulus (selective perception), providing learning guidance (semantic encoding), eliciting performance (responding), providing feedback (reinforcement), assessing performance (retrieval), and enhancing retention and transfer (generalization) (Aronson et al., 1983).

45

For Gagne, mathematics is composed of a set of tasks to be learned and occurs hierarchically. The learning task is analyzed to their subordinate elements, and mastery of each subordinate element is essential to the attainment of the main task. That is, learning of the task cannot occur without mastering their subordinate elements and therefore mathematics instruction must start with the subordinate elements of the task.

2.2.3 Landa and the Learning / Teaching of Mathematics Landa’s Algo-Heuristic theory (1976) is a general theory of learning, but it is illustrated primarily in the context of mathematics and foreign language instruction. It is concerned with identifying mental processes that underlie expert learning, thinking and performance in any area. His theory represents a system of techniques for getting inside the mind of expert students and performers that enable one to uncover the processes involved. Once uncovered, they are broken down into their relative elementary components. Performing a task or solving a problem always requires a certain system of elementary knowledge units and operations. According to Landa, there are classes of problems for which it is necessary to execute operations in a well-structured, predefined sequence (algorithmic problems). For such problem classes, it is possible to formulate a set of precise unambiguous instructions (algorithms) as to what one should do mentally and / or physically in order to successfully solve any problem belonging to that class. There are also classes of problems (creative or heuristic problems) for which precise and unambiguous sets of instructions cannot be formulated. For such classes of problems, it is possible to formulate instructions that contain a certain degree of uncertainty (heuristics).

46

For the content of learning, Landa maintains that students have to learn not only knowledge but also the algorithms and heuristics of experts as well. Students also have to learn how to discover algorithms and heuristics on their own. Landa concentrates on the learning of cognitive operations, algorithms, and heuristics that make up general methods of thinking (i.e., intelligence). According to this point of view, Landa affirms that learning algo-heuristic processes is more important than learning prescriptions (knowledge of processes). For Landa, mathematics is also composed of a set of specific and general problems or tasks that can be identified and taught sequentially. In this manner, Landa proposes the “snowball” method of learning / teaching. This method entails the following sequence: The first elementary operation in the chain is taught / learnt and practiced alone, then the second elementary operation is taught / learnt alone, practiced alone then is practiced together with the first, then the third is taught / learnt alone, practiced alone and then practiced together with the first two, and so on, until all elementary operations have been taught separately but practiced together (Landa, 1983). 2.2.4 Scandura and the Learning / Teaching of Mathematics According to Scandura’s Structural theory (1980), learning occurs through learning rules that consist of domain (internal cognitive structures of relevant environmental elements of a learning situation), range (expected rule that corresponds to the cognitive structure a student utilize to complete an objective), and procedures or operations (sum of all decisions and actions to produce a specific range element). Learning starts by using a very simple task as a prototype. Doing so requires identifying the educational goals first and then identifying prototypic cognitive processes (rules). There may be alternative rule sets for any given class of tasks.

47

Scandura identifies two types of rules: higher order rules and lower order rules. Higher order rules generate new rules, and problem solving may be facilitated when higher order rules are used. Higher order rules account for creative behavior (unanticipated outcomes) as well as the ability to solve complex problems by making it possible to generate (learn) new rules. Lower order rules are simple rules that the learning task starts with and are later reduced in number to derive higher order rules, the redundant lower order rules from the rule set will be eliminated by the student after practice and mastery that task. The rules which are to be learned can be derived from educational goals through structural analysis which is a methodology for identifying the rules to be learned for a given topic or class of tasks and breaking them down into their atomic components. The major steps in structural analysis are: (1) select a representative sample of problems, (2) identify a solution rule for each problem, (3) convert each solution rule into a higher order problem whose solutions is that rule, (4) identify a higher order solution rule for solving the new problems, (5) eliminate redundant solution rules from the rule set, and (6) continue the process iteratively with each newly-identified set of solution rules. The result of repeatedly identifying higher order rules, and eliminating redundant rules, is a succession of rule sets, each consisting of rules which are simpler individually but collectively more powerful than the ones before. Structural theory suggests that instruction has to start with the simplest solution path for a problem and then move to the more complex paths until the student masters the entire rule. The theory proposes that higher-order rules should be taught through elaboration and replacement of lower order rules. The theory also suggests a strategy

48

for individualizing instruction by analyzing which rules a student has / has not mastered and teach only the rules, or portions thereof, that have not been mastered. Structural theory has been applied extensively to mathematics. Here is an example in the context of subtraction provided by Scandura (1977):

1. The first step involves selecting a representative sample of problems such as 9-5, 248-13, or 801-302. 2. The second step is identifying the rules for solving each of the selected problems. This step can be achieved by determining the minimal capabilities of the students (e.g., can recognize the digits 0-9, minus sign, column and rows). Then the detailed operations involved in solving each of the representative problems must be worked out in terms of the minimum capabilities of the students. For example, one subtraction rule students might learn is the borrowing procedure that specifies if the top number is less than the bottom number in a column, the top number in the column to the left must be made smaller by 1. 3. The next step is identifying any higher order rules and eliminating any lower order rules they subsume. In the case of subtraction, a number of partial rules can be replaced with a single rule for borrowing that covers all cases. 4. The last step is to testing and refining the resulting rules by applying to new problems and extending the rule set if necessary so that it accounts for all problems in the domain. In the case of subtraction, problems with varying combinations of columns may be used.

49

Scandura also sees mathematics as a set of existing rules or procedures and the goal of teaching is to develop expertise by deriving general rules from specific rules. Therefore, teaching process starts with the simplest solution path first and then moves to the more complex paths or rule sets, that is, the appropriate sequence in teaching is from bottom up. The students’ prior capabilities have to be taken into account before teaching rules. So rules must be composed of the minimum capabilities possessed by the students. The objectivist theories and models are based on the view that knowledge of the world is fixed and can be quantified. These theories and models call for information or knowledge to be taught to be divided into parts that are slowly assembled into whole concepts. Mathematics according to these theories and models is seen as a set of preexisting facts and procedures, free of context and value. Mathematics knowledge is passed along from those who know to those who do not through authoritative means, including memorization and practice (Borasi, 1996). So teachers serve as pipelines or assemblers of knowledge and seek to transfer their thoughts and meanings to the students. Lessons derived from these theories and models are teacher-centered and depend heavily on textbooks for the structure of the course. The students are generally passive or compliant, and there is little room for student-initiated questions, independent thought or interaction between them. The role of the students is to regurgitate the accepted explanation or methodology expostulated by the teacher (Hanley, 1994). Also the assessment of performance is end-centered, that is, the concentration is on mastering the task. Being content-based, these theories and models produce lessons that are presented below the student’s true cognitive ability.

50

These objectivist theories and models of teaching do not meet the needs of learning mathematics with understanding where the student is actively doing mathematics through the process of inquiry and investigation (von Glasersfeld, 1995). These theories and models also do not promote mathematical reasoning, adaptive expertise and an ability to deal with change or solve ill-structured problems characteristic of today’s complex society (NCTM, 2000). 2.3 Constructivist Views Regarding the Learning / Teaching of Mathematics 2.3.1 Nature of the Learning Process and Construction of Knowledge The central idea of objectivist is that learning performance could be defined solely in terms of observable behavior, and the teacher’s job is just to give orders and monitor student responses. New theories soon emerged to challenge the behaviorists, the earliest being Gestaltism (Schoenfeld, 1987). Gestalt theory is one of the early learning theories which emphasize the role of understanding. It is also one of the first to deal with issues of problem solving and creativity. Wertheimer (1959) is one of the principal proponents of Gestalt theory that emphasizes higher-order cognitive processes. The focus of Gestalt theory is the idea of grouping, i.e. characteristics of stimuli cause us to structure or interpret a visual field or problem in a certain way. The primary factors that determine grouping were: (1) proximity - elements tend to be grouped together according to their nearness, (2) similarity - items similar in some respect tend to be grouped together, (3) closure - items are grouped together if they tend to complete some entity, and (4) simplicity - items will be organized into simple figures according to symmetry, regularity, and smoothness. These factors are called the laws of organization and are explained in the context of perception and problem solving.

51

For learning mathematics, the essence of successful problem-solving behavior according to Wertheimer (1959) is being able to see the overall structure of the problem: "A certain region in the field becomes crucial, is focused; but it does not become isolated. A new, deeper structural view of the situation develops, involving changes in functional meaning, the grouping, etc. of the items. Directed by what is required by the structure of a situation for a crucial region, one is led to a reasonable prediction, which like the other parts of the structure, calls for verification, direct or indirect. Two directions are involved: getting a whole consistent picture, and seeing what the structure of the whole requires for the parts." (P 212). How humans learn has intrigued and troubled educators throughout history. Constructivists believe that learning involves the generation of knowledge and learning strategies. According to this view, learning in schools has to emphasize the use of intentional processes that students can use to construct meaning from information, experiences, and their own thoughts and beliefs. Mayer (1996) asserts that successful students are active, goal-directed, self-regulating, and assume personal responsibility for contributing to their own learning. So the learning of complex subject matter is most effective when it is an intentional process of constructing meaning from information and experience. von Glasersfeld (1995) argues that: “From the constructivist perspective, learning is not a stimulus-response phenomenon. It requires self-regulation and the building of conceptual structures through reflection and abstraction” (p.14). Fosnot (1996) mentions that “Rather than behaviors or skills as the goal of instruction, concept development and deep understanding are the foci” (p.10). This view of learning sharply contrasts with the one in which learning is the passive transmission of information from one individual to another. The psychological theoretical base for constructivism comes from Piaget. He uses the term schemata to describe mental or cognitive structures that allow one to think about,

52

organize and make sense of experiences (Borich and Tombari, 1997). The individual continuously constructs his or her schemata. Cognitive development, or learning, is the lifelong process by which the student constructs and modifies his or hers own personal schemata. This occurs in two ways: a) existing schemata are organized into higher-order, more complex structures, and b) the individual adapts to new information and experience through assimilating it or accommodating it. Consequently, the constructivist point of view defines meaning as an act of interpretation i.e., meaning does not exist independently of the student (Mugny and Doise, 1978). For mathematics, White (1998) maintains that “Educational research has shown that students tend to comprehend complex concepts much better and to retain them as part of their body of knowledge much longer when they become actively involved in their learning process” (p.1). Mathematics can be actively learned by involving students in their leaning process. Ahmed (1987) asserts that “Mathematics can be effectively learned only by involving students in experimenting, questioning, reflecting, discovering, inventing and discussing. Mathematics should be a kind of learning which requires a minimum of factual knowledge and a great deal of experience in dealing with situations using particular kinds of thinking skills” (p.24). Carpenter and Lehrer (1999) indicate that the critical learning of mathematics by students occurs as a consequence of building on prior knowledge via purposeful engagement in activities and by discourse with other students and teachers in classrooms. So students must engage in activities that encourage their mathematical understanding. This view of learning mathematics leads to the characteristics of learning mathematics with understanding. Hiebert and Carpenter (1992) assert that learning mathematics

53

with understanding implies students not only must learn the concepts and procedures of mathematics (its design features), but they must learn to use such ideas to solve non-routine problems, and learn to mathematize in a variety of situations (its social functions). Therefore the concentration should shift from judging student learning in terms of mastery of concepts and procedures to making judgments about students’ deep understanding of the concepts and procedures and their ability to apply them to mathematics problem situations. The construction of relationships is one of the important forms of mental activity where mathematical understanding emerges. For students to learn mathematics with understanding, new ideas take on meaning by the ways they are related to other ideas. Students construct meaning for a new idea or process by relating it to ideas or processes that they have already understood. Although learning with understanding entails forging connections between what the students already know and the knowledge they are learning, it is not sufficient to think of developing understanding simply as appending new concepts and processes to existing knowledge. Over the long run, developing understanding involves more than simply connecting new knowledge to prior knowledge; it also involves the creation of rich, integrated knowledge structure (Carpenter and Lehrer, 1999). Therefore, the role of the teacher and the role of students should be appropriate to this learning environment. The teacher’s role should be shifted from being an orator to a learning manager and facilitator who manages, directs, and encourages students’ creation or from sage on the stage to guides on the side where he provides students with opportunities to test the adequacy of their current understandings. Doyle (1988) argues that teachers should be especially attentive to the extent to which meaning is

54

emphasized and the extent to which students are explicitly expected to demonstrate understanding of the mathematics underlying the activities in which they are engaged. Such an emphasis can be maintained if explicit connections between the mathematical ideas and the activities in which students engage in are frequently drawn. Also, Carpenter and Lehere (1999) assert that connections with what students already know and understand what they are learning play an important role in engaging students in high-level thought processes. The mathematical activities should therefore be selected to encourage the students to link between the knowledge what they have already learned and the new knowledge. The students’ role should also be changed from obtaining knowledge from the teacher to assimilating or accommodating their own knowledge by connecting the relationships between what they have known and what they are learning. The students’ role should also be shifted to confront their understanding in light of what they encounter in the new learning situation (Manion, 1995). If what the students encounter is inconsistent with their current understanding, their understanding can change to accommodate new experience. The teacher has to keep in mind that students come to the learning situations with knowledge gained from previous experience, and that prior knowledge influences what new or modified knowledge they will construct from the new learning experiences. Bennett and Desforges (1988) affirm that a critical factor underlying unsuccessful task implementation is a lack of alignment between tasks and students' prior knowledge, interests, and motivation. Such mismatches may cause students to fail to engage with the task in ways that will maintain a high level of cognitive activity. 2.3.2 Piaget and the Learning / Teaching of Mathematics

55

Cognitive structure is the central idea of Piaget’s (1970) theory. Cognitive structures are patterns of physical or mental action that underlie specific acts of intelligence and correspond to stages of child development. Piaget specifies four primary cognitive structures (i.e., development stages): sensorimotor (0-2 years) where intelligence takes the form of motor actions; preoperations period (3-7 years) where intelligence is intuitive in nature; The concrete operational stage (8-11 years) where the cognitive structure is logical but depends upon concrete referents; and formal operations (12 and above) where thinking involves abstractions. While the stages of cognitive development identified by Piaget are associated with characteristic age spans, they vary from one individual to another. Piaget indicates that cognitive structures are not stable and they change through the processes of adaptation i.e., assimilation and accommodation. Assimilation involves the interpretation of events in terms of existing cognitive structure whereas accommodation refers to changing the cognitive structure to make sense of the environment. Cognitive development consists of a constant effort to adapt to the environment in terms of assimilation and accommodation. Piaget affirms that all children construct, or create logic and number concepts from within rather than learn them by internalization from the environment (Piaget 1971; Piaget and Szeminska 1965; Inhelder and Piaget 1964; and Kamii 2000). Piaget explores the implications of his theory to all aspects of cognition, intelligence and moral development. Many of Piaget’s experiments were focused on the development of mathematical and logical concepts. Applying Piaget’s theory, results in specific recommendations for a given stage of cognitive development. For example, with students in the concrete operational stage, learning activities should involve

56

problems of classification, ordering, location, and conservation using concrete objects. So teachers should also try to provide a rich and stimulating environment with logical matters that depend on concrete objects and try to prepare students to the next stage, namely, formal operational that involves abstract thinking.

2.3.3 Vygotsky and the Learning / Teaching of Mathematics A critical factor that relates to the learning process and construction of knowledge is sociocultural development. Constructivists view sociocultural development as one of the significant factors that contribute to the construction of knowledge. Vygotsky (1978) states that cognitive development is dependent on social interaction, and that cultural development has two levels: social and interpersonal. During social interaction, the students recognize the new knowledge and then internalize it. So for effective learning, students have to cooperate in an environment where social interaction is taken into account (Bonk and Reynolds, 1997). Vygotsky suggests that students can be guided by explanation, demonstration, and can attain to higher levels of thinking if they are guided by more capable and competent adults. This conception is better known as the Zone of Proximal Development (ZPD). The Zone of Proximal Development is the gap between what is known and what is not known, that is, generally higher levels of knowing. The ability to attain higher levels of knowing is often facilitated and, in fact, depends upon, interaction with other more advanced peers, who for Vygotsky are generally adults. Through increased interaction and involvement, students are able to extend themselves to higher levels of cognition. Vygotsky defines the Zone of Proximal Development as “the distance between the actual developmental level as determined by independent problem solving and the

57

level of potential development as determined through problem solving under the guidance or in collaboration with more capable peers.” The ZPD is thus the difference between what students can accomplish independently and what they can achieve in conjunction or in cooperation with another, more competent person. The sociocultural development enriches the active learning processes and contributes in encouraging constructing knowledge. Moore and Kearsley (1996) have indicated that sociocultural development is an area that is missing in most traditional or objectivist learning environments. Interactions between the student and the content, between the student and the instructor, and between the students themselves are necessary for learning and for the shared social construction of knowledge (American Psychological Association [APA], 1995; Moore and Kearsley, 1996).

2.3.4 Bruner and the Learning / Teaching of Mathematics A major theme in the theoretical framework of Bruner (1960) is that learning is an active process in which, students construct new ideas or concepts based upon their current or past knowledge. The student selects and transforms information, constructs hypotheses, and makes decisions, relying on a cognitive structure to do so. Cognitive structure i.e., schema, mental models provides meaning and organization to experiences and allows the individual to go beyond the information given. Bruner believes that students can and have to discover knowledge by themselves. So the teacher should encourage students to discover their knowledge. The teacher and student should engage in an active dialog i.e., Socratic learning where the teacher’s task is to translate information to be learned into a format appropriate to the students’ current state of understanding. To reach that learning environment, Bruner suggests

58

that the curriculum should be organized in a spiral manner so that the students continually build knowledge upon what they have already learned. He maintains that learning starts from the top down, that is, it begins with problem solving that makes students learn the fundamentals because they need them. Instruction for him is a roller coaster ride of successive disequilibrium and equilibrium until the desired cognitive state is reached or discovered (Shulman, 1973). Bruner asserts that knowing is a process not a product. He describes three levels of student’s representation: the enactive level where the student manipulates materials directly. The second level is the ikonic level where the student deals with mental images of objects but does not manipulate them directly. The final level is the symbolic level where the student is strictly manipulating symbols and no longer mental images of object. This sequence is an out growth of the developmental work of Piaget (Shulman, 1973). Transfer of learning for Bruner (1960) occurs when the student can identifying from the structures of subject matters basic, fundamentally simple concepts or principles which, if learned well, can be transferred both to other subject matters within that discipline and to other disciplines as well. He gives an example of the concept of balance. If the teacher teaches the balance of trade in economics in such a way that when ecological balance is considered, students will see the parallel and this could be extended to balance of power in political science, or to balancing equations. In summary, learning according to Bruner is a process, and through the processes, students discover and build their knowledge. Instruction should be concerned with the experiences and contexts that make the student willing and able to learn (readiness), it should be structured so that it can be easily grasped by the student (spiral

59

organization), and instruction should be designed to facilitate extrapolation (going beyond the information given) (Bruner, 1973).

2.4 Metacognitive Strategies and the Construction of Knowledge The central point of constructivism is that learning involves more than just the transfer of information from the teacher to a student; instead, each student plays an active role in working with and integrating the information according to his or her own background or experience. This integration involves applying personal study and learning skills, and monitoring one’s own comprehension (Gordon, 1996). Therefore, to construct or reconstruct his or her knowledge, the student needs to employ certain techniques regarding managing his or her thinking like thinking about thinking, planning, monitoring, and evaluation. In other words for students to reach the equilibrium case (resolving the conflicts), and then assimilate or / and accommodate their knowledge, they should employ various metacognitive strategies. Metacognitive strategies are techniques that students use to plan, monitor and control, and evaluate their own cognitive processes (Flavell, 1976). These strategies are seldom taught directly and tend to develop naturally in only good students (Smith and Ragan, 1993). However Jacobson (1998), Perkins and Grotzer (1997), and Halpern (1996) indicate that metacognitive strategies can be systematically taught to most students. Also Paris and Winograd (1990) argue that teachers can promote metacognitive strategies directly by guiding students about effective problem solving strategies and discussing cognitive and motivational characteristics of thinking.

60

Metacognition has been defined as an awareness of one’s own cognitive processes rather than the content of those processes together with the use of that self-awareness in controlling and improving cognitive processes (Biggs and Moore, 1993). Other researchers have referred to metacognition as “cognitive strategies” (Paris and Winograd, 1990), “knowledge about executive control systems” (Brown et al., 1994), “monitoring of cognitive processes” (Flavell, 1976), “resources and self regulating learning” (Osman and Hannafin, 1994, Lawson, 1995) and “knowledge and regulation of cognition, and evaluating cognitive states such as self-appraisal and selfmanagement” (Brown, 1987). Flavell (1977) refers to metacognition as metamemory which he describes as intelligent structuring and storage, intelligent search and retrieval, and intelligent monitoring. His description suggests that metacognitive strategies are deliberate, planful, intentional, goal-directed, and future-oriented mental behaviors that can be used to accomplish cognitive tasks. Flavell mentions that metacognition is an awareness of oneself as “an actor in his environment, that is, a heightened sense of the ego as an active, deliberate storer and retriever of information” (p. 275). “It is the development of memory as applied cognition” (p. 273), in which whatever “intellectual weaponry the individual has so far developed” is applied to mnemonic problems (p. 191). What is basic to the concept of metacognition is the notion of thinking about one’s own thoughts. These thoughts can be of what one knows (i.e., metacognitive knowledge), what one is currently doing (i.e., metacognitive skill), or what one’s current cognitive or affective state is (i.e., metacognitive experience). To differentiate metacognitive thinking from other kinds of thinking, it is necessary to consider the

61

source of metacognitive thoughts: Metacognitive thoughts do not spring from a person’s immediate external reality; rather, their source is tied to the person’s own internal mental representations of that reality, which can include what one knows about that internal representation, how it works, and how one feels about it. Therefore, metacognition sometimes has been defined simply as thinking about thinking, cognition of cognition, or using Flavell’s (1979) words, “knowledge and cognition about cognitive phenomena” (Hacker,1998, p. 906). Although perspectives differ in emphasis, there is common agreement that metacognitive strategies involve both the knowledge of and the regulation of cognition. Pressley and McCormick (1987) identify two components of knowledge of cognition, which are knowledge about and awareness of one’s own thinking and knowledge of when and where to use acquired strategies. Knowledge about one’s thinking includes information about one’s own capacities and limitations and awareness of difficulties as they arise during learning so that remedial action may be taken. Knowledge of when and where to use acquired strategies, includes knowledge about the task and situations for which particular goal-specific strategies are appropriate. In the absence of domain-specific knowledge or lack of information in various content areas, students often need to apply general strategies, which can be applied to the problems, regardless of their content. In the social science study conducted by Voss et al. (1991), they found that experts were flexible in that they take into account more factors than do novices in searching for information. Additionally, experts used strategies of argumentation more often than novices did. They concluded that argumentation may be an important strategy in problem solving (Gick, 1986).

62

2.4.1 Regulation of Cognition Jacobs and Paris (1987) determine three components of regulation of cognition, which are planning, monitoring, and evaluation. Planning (processes selected prior to any task action) consists of setting goals, activating relevant resources, and selecting appropriate strategies. Monitoring (processes selected to keep track of what has been done, what is currently being done, and what still needs to be done for task solution) involves checking one’s progress and selecting appropriate repair strategies when originally-selected strategies are not working. Evaluation (processes selected to judge the outcome of any action against criteria of effectiveness and efficiency, evaluation refers to students’ ongoing assessments of their knowledge or understanding, resources, tasks, and goals) involves determining one’s level of understanding. In short, regulation of cognition is thinking before, during and after a learning task. Jacobs and Paris (1987); and North Central Regional Educational Laboratory, (NCREL, 1995) point out that successful students ask themselves metacognitive questions before (through planning), during (through monitoring), and after (through evaluation) the learning task. For example, at the planning stage the student asks him or herself metacognitive questions such as: “What in my prior knowledge will help me with this particular task? What should I do first? Do I know where I can go to get some information on this topic? How much time will I need to learn this? What are some strategies that I can use to learn this?” At the monitoring stage the successful student asks him or herself metacognitive questions such as: “Did I understand what I just heard, read or saw? Am I on the right track? How can I spot an error if I make one? How should I revise my plan if it is not working? Am I keeping good notes or records?” And at the evaluation stage the student asks him or herself metacognitive

63

questions such as: “Did my particular strategy produce what I had expected? What could I have done differently? How might I apply this line of thinking to other problems?”

2.4.2 Metacognitive Strategies and Age The idea of deliberate, planful, and goal-directed thinking applied to one’s thoughts to accomplish cognitive tasks is deeply embedded in Piaget’s conceptualization of formal operations in which higher-ordered levels of thought operate on lower-ordered levels. During this stage of cognitive development, the abilities of the adolescent begin to differentiate from those of the child (Hacker,1998). Flavell (1963) wrote: “What is really achieved in the 7-11-year period is the organized cognition of concrete objects and events per se (i.e., putting them into classes, seriating them, setting them into correspondence, etc.). The adolescent performs these first-order operations, too, but he does something else besides, a necessary something which is precisely what renders his thought formal rather than concrete. He takes the results of these concrete operations, casts them in the form of propositions, and then proceeds to operate further upon them, i.e., make various kinds of logical connections between them (implications, conjunction, identity, disjunction, etc.). Formal operations, then, are really operations performed upon the results of prior (concrete) operations. Piaget has this propositions-about-propositions attribute in mind when he refers to formal operations as second-degree operations or operations to the second power” (p. 205206). Inhelder and Piaget (1958) provide further elaboration on second-degree operations: “... this notion of second-degree operations also expresses the general characteristics of formal thought. It goes beyond the framework of transformations bearing directly on empirical reality (first degree operations) and subordinates it to a system of hypothetico-deductive operations--i.e., operations which are possible” (p. 254). Thus, first-degree operations, which are thoughts about an external empirical reality, can become the object of higher-order thoughts in an attempt to discover not necessarily

64

what is real but what is possible. “Formal thinking is both thinking about thought... and a reversal of relations between what is real and what is possible” (p. 341-342). Referring to Inhelder and Piaget’s work, Flavell (1977) wrote: “Another way to conceptualize it is to say that formal operations constitute a kind of metathinking, i.e., thinking about thinking itself rather than about objects of thinking. Children certainly are not wholly incapable of this and other forms of metacognition” (p. 107). So for young students, 11 year olds for example, they can think metacognitively and apply metacognitive strategies in their learning processes, particularly if they are taught and supported deliberately how and when to use these strategies. 2.4.3 Metacognitive Scaffolding Students in 7-11 years stage (concrete operations) have some abilities, and some higher levels of thinking that enable them to work in the next stage (formal operations), but they need a certain guidance and support from more capable and competent adults to reach that stage. These children need to narrow their zone of proximal development; they can be pushed to the next stage or can narrow their ZPD by scaffolding and supporting them. Vygotsky (1978) believes that students cannot independently narrow the zone of proximal development (Rosenshine and Meister, 1993). So the concept of scaffolding becomes a critical technique to bridge the gap between what the students can accomplish independently and what they can achieve with assistance or guidance of others. Therefore, scaffolding is a technique of teaching where the learning is assisted by the teacher or / and other capable peers (Slavin, 1994; Rosenshine and Meister, 1993). When using scaffolding, students are provided with “a great deal of support during the early stage of learning and then diminishing support and having the students take on increasing responsibility as soon as they are

65

able” (Slavin, 1994, p. 49). In this way, students are able to narrow the zone of proximal development initially with support, and retain this level of achievement as support is reduced. So awareness of a student’s ZPD helps a teacher gauge the tasks student is ready for, the kind of performance to expect, and the kinds of tasks that will help the student reaching his or her potential. Fading of support during scaffolding should eventually result in self-regulated learning, and thus more self-reliant students. Recent developments in pedagogy and educational science also picture this more active, self-reliant role of students, selfregulating their own learning process and actively creating new knowledge. For selfdirected learning, metacognition, “one’s awareness of one’s own cognition” (Alessi and Trollip, 2001, p. 28), is needed which is so helpful for life long learning. As students are being supported to work self-reliantly, they can learn how to learn, which is critical for their professional futures where they will be required to keep themselves up-to-date in their own professions. Brown et al. (1991) describe scaffolding in reciprocal teaching which enhances interactive learning. Interactive learning provides students with situations that push the boundaries of their abilities and actively engage them in tasks. It also gives students an opportunity to be students as they come to master a task and, once they have achieved mastery, to be teachers of those who are still learning. Brown et al. (1991) add that research indicates that problems which are too difficult at first for students to handle on their own, later become problem types they can solve independently when they have first received support and worked on them in a small group setting. That is, the teacher scaffolds students and students scaffold themselves. Therefore, scaffolding enables students to learn a body of coherent, usable, and meaningful knowledge within their zone of proximal development and “to develop a

66

repertoire of strategies that will enable them to learn new content on their own” (p. 150). King (1991a) affirms that matacognitive scaffolding, in the form of metacognitive questions help students to clarify the problem and access their existing knowledge and strategies when relevant. For example, to identify or redefine the problem, questions such as, “What are you trying to do here?” can be asked which is expected to help students determine the nature of the problem more precisely. Questions such as “What information is given to you?” would presumably help students to access prior knowledge, whereas the question “Is there another way to do this?” would foster greater access to known strategies. A question to monitor problem solving may be “Are you getting close to your goal?” Evaluation questions such as, “ Does the solution make a sense?” enable students to reflect on their problem solving process, for instance, to articulate the steps they have taken and decisions they have made, facilitating their understanding of the reasons behind actions. In sum, metacognitive scaffolding guide students’ attention to specific aspects of their learning process, helping students to plan, monitor, and evaluate their own learning processes, and therefore, helping students to learn with understanding (Lin et al., 1999). King (1991a) found that many students lack the ability to engage in effective thinking and problem solving on their own; therefore, scaffolding in the form of metacognitive questions should be made to support students to plan, monitor, and evaluate their learning, and therefore, learn with understanding. Moreover, this scaffolding is likely to enable students to make judgments about what can be known and what cannot and to justify the problem solution. Questions such as “What is your justification for that

67

solution?” would help students to construct cogent arguments for their point of view (Jonassen, 1997). Kramarskis et al. (2001) found that students with metacognitive scaffolding ask themselves questions about the nature of the problem (what is the problem all about?), about the appropriate strategies to solve the problem (what are the appropriate strategies to solve the problem, and why?), and about the construction of relationships between the previous and the new knowledge (What are the similarities / differences between the problem at hand and the problems solved in the past?). So students who are metacognitively scaffolded will more than likely be students who plan, monitor, and evaluate their learning processes and outcomes. In other words, they will more than likely be able to refer to the what, how, when, where and why of the learning processes and solutions. Wong (1985) affirms that teaching students to ask questions help them become sensitive to important points in a text and thus monitor the state of their reading comprehension. Palincsar and Brown (1984) indicate that in asking and answering questions concerning the key points of a selection, students are likely to find that problems of inadequate or incomplete comprehension can be identified and resolved. Van Zee and Minstrell’s (1997) study described “a reflective toss” through a questionanswer cycle between the teacher and the students, which revealed the influence of a teacher’s questions on a student’s reflective thinking process. It is evident that metacognitive questions serve to facilitate metacognition in planning by activating prior knowledge and attending to important information, in monitoring by actively engaging students in their learning process, and in evaluation through reflective thinking.

68

Therefore, in the present study, the teacher in the cooperative learning with metacognitive scaffolding group scaffolded students by asking metacognitive questions and students were coached to ask themselves and their group members metacognitive questions based on materials presented in the classroom. When students in cooperative learning settings ask and answer the metacognitive questions, they are more likely to understand the materials better, develop new perspectives, reason and explain solutions, and recognize and fill in gaps in their understanding. 2.5 Cooperative Learning and Learning Mathematics with Understanding A common response to the idea that students construct their knowledge is that students should be encouraged to work with others. As Dewey (1961) and Vygotsky (1978) suggest, people do not learn in isolation from others, they naturally learn and work cooperatively throughout their lives (Petraglia, 1998). Cooperative learning is defined differently by different researchers and theorists. Vygotsky (1978), for example, views cooperative learning as part of a process leading to the social construction of knowledge. Other scholars (Kohn, 1992; Sapon-Shevin and Schniedewand, 1992) consider cooperative learning to be a form of critical pedagogy that moves classrooms and societies closer toward the ideal of social justice. Caplow and Kardash (1995) characterize cooperative learning as a process in which “knowledge is not transferred from expert to student, but created and located in the learning environment” (p.209). Others such as Burron et al. (1993) and Ossont (1993) see cooperative learning as a strategy to help students improving intellectual and social skills.

69

Although there seems to be some differences between the definitions of cooperative learning, there is common agreement that cooperative learning is an instructional method in which small groups of students work together to accomplish a shared goal through changing or reconstructing their knowledge. The aim of this cooperation is for students to maximize their own and each other’s learning, with members striving for joint benefit. 2.5.1 Theoretical Perspective on Cooperative Learning Piaget (1970) focuses on the individual as a starting point. Knowledge or information is provided through cooperation for the individual to use when becoming aware of differing perspectives and in resolving the differences between them. Cognitive development from Piagetian view is the product of an individual, perhaps sparked by having to account for differences in perspectives with others (Rogoff, 1990). The process of knowing for Piaget comes about through the sequence of equilibrium, disequilibrium, and reequilibrium. In these processes, existing schemes are altered to accommodate new information or new information is being assimilated into existing schemes, which are then strengthened. Piaget stresses that these processes can occur either by way of cognitive conflicts, in which intraindividual discord during thinking / problem solving leads to reequilibrium, or by way of sociocognitive conflicts, in which intreindividual differences during thinking / problem solving are catalysts for cognitive growth (Manion, 1995). Piaget believes that individuals work with independence and equality on each other’s ideas, so when they interact they learn, receive feedback, or are told of something that contradicts their believes or current understanding. This is what Piaget calls “cognitive conflicts” (Mugny and Doise, 1978). This conflict initiates a process of cognitive or intellectual reconstruction in an individual. Therefore, students’ interaction prompts the student to assimilate or

70

accommodate his or her knowledge. As Manion (1995) indicates, students revise their ways of thinking to provide a better fit with reality when faced with discrepancies between their own ways of viewing the world and new information. Vygotsky is another psychologist who has done extensive work in social context. In contrast to Piaget, Vygotsky (1978) focuses on the social basis of mind. He believes that an individual makes use of the joint decision-making process itself to expand understanding and skill. Cognitive development involves an individual’s appropriation or internalization of the social process as it is carried out externally in joint problem solving. Vygotsky (1978) affirms that individual intellectual development cannot be understood without reference to the social setting in which the student is embedded. Students’ social interaction with more competent students is essential to cognitive development. So students’ cognitive or learning is developed through interaction with more skilled partners working in the zone of proximal development. This interaction enables students to discuss and exchange their ideas and thoughts which in turn emulate rational thinking processes such as the verification of ideas, the planning of strategies in advance, the symbolic representation of intelligent acts, spontaneous generation, and criticism. Student will then takes on and internalize these procedures thus enhancing the development of his or her intellectual abilities such as his or her problem solving capacities. Although there seems to be some differences between Piaget and Vygotsky, Ismail (1999) believes that they actually complement each other, and if they are combined, they provide a profound conceptual base for cooperative learning.

71

Cooperative learning is the instructional use of small groups so that students work together to maximize their own and each other’s learning. Students perceive that they can reach their learning goals if and only if the other students in the learning group also reach their goals (Johnson, 1981). They are not only responsible for learning the material that is presented, but also for ensuring everyone in the group knows the material as well (Slavin, 1987). So students need not only to interact with materials (i.e., textbooks, curriculum programs) or with the teacher, but also they need to interact with each other to achieve their learning goals. Johnson and Johnson (1999) identify three basic types of learning that goes on in any classroom: competitive learning where students compete to see who is the best, individualistic learning where students work individualistically toward a goal without paying attention to other students, and cooperative learning where students work cooperatively with a vested interest in each other’s learning as well as their own. Of the three patterns, competition is presently the most dominant where the students view the school as a competitive enterprise where one tries to do better than others. Cooperation among students who celebrate each other’s successes, encourage each other to do homework, and learn to work together regardless of ethnic backgrounds or whether they are male or female, bright or struggling, disabled or not, is still rare. Even though these three patterns are not equally effective in helping students learn, it is important that students learn to interact effectively in each of these ways. Students will face situations in which all three patterns are operating and they will need to be able to be effective in each. They also should be able to select the appropriate pattern suited to the situation. However, in the ideal classroom all three patterns are used. This does not mean that they should be used equally, but the cooperative pattern should

72

dominate the classroom, being used 60 to 70 percent of the time. The individualistic pattern may be used 20 percent of the time, and a competitive pattern may be used 10 to 20 percent of the time (Johnson and Johnson, 1999). According to Johnson et al. (1986), about 600 experimental and over 100 correlational studies have been conducted since 1898 which have compared these three learning types or patterns. The majority of the studies show that cooperative learning has advantages over the competitive and individualistic learning. Moreover, many learning and instructional approaches that apply cooperative learning have resulted in the students’ cognitive, intellectual, social, and affective growth (Johnson et al., 1991; Slavin, 1996). Johnson and Johnson (1999) clarify that there is a difference between simply having students work in a group and structuring groups of students to work cooperatively. A group of students sitting at the same table doing their own work, but free to talk with each other as they work, is not structured to be a cooperative group, perhaps it could be called individualistic learning with talking. For this to be a cooperative learning situation, there needs to be an accepted common goal on which the group is rewarded for its efforts. If a group of students has been assigned to do a report, but only one student does all the work and the others go along for a free ride, it is not a cooperative group. A cooperative group has a sense of individual accountability that means that all students need to know the material well for the whole group to be successful. In other words, a group of students can be a cooperative learning group if the elements of cooperative learning are fulfilled. 2.5.2 Elements of Cooperative Learning

73

Johnson and Johnson (1987) have identified five basic elements of cooperative learning. These include: Promotive, Face to Face Oral Communication: Students are placed in heterogeneous groups from 2 to 6 members. Team members are strategically seated in order to encourage “eye-to-eye, knee-to-knee” interaction. Through team building activities, promotive behavior is facilitated. Positive interdependence: “All for one and one for all” “ Sink or swim together”. As students work toward a common goal, team cooperation and fellow success becomes imperative. Individual accountability: What students can do together today, they can do alone tomorrow (Vygotsky, 1978). Although students work together in a cooperative group, each student is held accountable for individual learning. Individual student performance is assessed and the outcome is reported and celebrated by the individual as well as team members. Interpersonal, Cooperative Social Skills: Students work together to reach a common goal. In order for members to reach a common goal, students must utilize adequate cooperative social skills to function successfully. Evaluating and processing: Students are given time and encouraged to participate in reflection about what was learned, how it was learned, and the skills used to process and meet the goal. 2.5.3 Teacher’s Role in Cooperative Learning

74

Through the discussion above, the role of the teacher in cooperative learning can be shown as a facilitator rather than prompter, a supervisor rather than instructor. The teacher specifies the instructional objectives, monitors the learning groups, asks questions, and intervenes when necessary. Also the teacher contributes in deciding the group size and assigning group members and roles (e.g., recorder, summarizer, encourager, checker, etc). Finally, the teacher contributes in refinement and evaluation processes of learning outcomes. That is, teacher’s role is to guide and support students to build or reconstruct their knowledge, to be a guide on the side rather than a sage on the stage. For learning mathematics, and according to the constructivist theories, information is retained and understood through elaboration and construction of connections between prior knowledge and new knowledge (Wittrock, 1986). Because providing explanations is one of the best means for elaborating information and making connections (Slavin, 1996) and students in cooperative settings often give explanations to each other, the likelihood of constructing rich networks of knowledge under these conditions increases. Also, when students work with peers who are in various stages of mastering a task, mutual reasoning and conflict resolution are likely to occur, which, in turn, facilitates learning (Mevarech and Light, 1992). Observing other students solving a problem help students internalize either the cognitive functions they are attempting to master or those that are within their zone of proximal development (Vygotsky, 1978). As students interact cooperatively, they explain strategies and mathematical ideas in their own words, thus helping one another to process complex cognitive activity (Schoenfeld, 1985). Researchers have emphasized the importance of mathematical communication to build students’ capacity for mathematical thinking and reasoning (Stein et al.,1996).

75

According to the National Council of Teachers of Mathematics (NCTM; 1991), learning environments should be created in the way that promote active learning and teaching; classroom discourse; and individual, small group, and whole-group learning. Cooperative learning is one example of an instructional arrangement that can be used to foster active student learning, which is an important dimension of mathematics learning and highly endorsed by math educators and researchers. Students can be given tasks to discuss, problem solve, and accomplish. Also Teachers can use cooperative learning activities to help students make connections between the concrete and abstract level of instruction through peer interactions and carefully designed activities. Finally, cooperative learning can be used to promote classroom discourse and oral language development. Wiig and Semel (1984) describe mathematics as “conceptually dense.” That is, students must understand the language and symbols of mathematics because contextual clues, like those found in reading, are lacking in mathematics. For example, math vocabulary (e.g., greater-than, denominator, equivalent) and mathematical symbols (e.g., =, <, or >) must be understood to work problems as there are no contextual clues to aid understanding. In a cooperative learning activity, vocabulary and symbolic understanding can be facilitated with peer interactions and modeling. In almost all studies, the metacognitive strategies were employed in cooperative learning settings where small groups of 4 – 6 students studied together (e.g., Schoenfeld, 1987; Mevarech and Kramarski, 1997; Hoek et al., 1999). The cooperative learning with metacognitive scaffolding method is based on cognitive theories of learning that emphasize the important role of elaboration in constructing

76

new knowledge (Wittrock, 1986), and on a large body of research (e.g., Davidson, 1990; Qin et al., 1995; Stacey and Kay, 1992; and Webb, 1991, 1989a) showing that cooperative learning has the potential to improve mathematics performance because it provides a natural setting for students to supply explanations and elaborate their reasoning. Since this potential has not always materialized, researchers suggested the embedding of metacognitive strategies in cooperative setting in order to provide an appropriate situation for students to elaborate their mathematical reasoning (Schoenfeld, 1992; Mevarech and Kramarski, 1997; Mevarech, 1999; Kramarski, 2000; and Kramarski et al., 2001). 2.6 Cooperative Learning with Metacognitive Scaffolding and Learning Mathematics with Understanding Most of people believe that mathematics in all is about computation. So most of them are familiar with only the computational aspects of mathematics and are likely to argue for its place in the school curriculum and for traditional methods of instructing students in computation. For them, the broad goal of learning process is to master the computational procedures regardless to what actually mathematics is about and regardless to the learning process itself. That is, they have misconceptions about what mathematics is about and they do not take how students learn, their experiences, their metacognitive strategies, and their attitudes toward mathematics into account (Brown, 1987; national Council of Teachers of Mathematics, 1989; Thompson, 1992). In most traditional mathematics classrooms, students are frequently expected to learn facts, concepts, and skills divorced from any real context. They are drilled in arithmetic without applying the skills to problems that mean anything to them. They usually learn abstract formulas in mathematics out of realistic contexts. So their

77

learning are likely ineffective, and although they can acquire mathematical operations they are usually unable to apply them in different situations as Clark (1995) has concluded that when students interpret an activity as unrealistic and non-meaningful, encoding, representation, and learning are likely to become over simplified and narrowly school-focused. Ertmer and Newby (1996) assert that “If schools are going to help all students become expert students, the metacognitive strategies of students must be acknowledged, cultivated, and exploited. A major function of all schooling must be to help create students who know how to learn” (p.22). Therefore, effective mathematics instruction should assist students to activate the metacognitive strategies in order to be able to learn mathematics with understanding and reason mathematically. Work in the area of mathematics problem solving suggests that the deployment of cooperative learning with metacognitive scaffolding underlie successful performance. Shoenfeld (1987) found that expert mathematicians engaged in decision-making and management behaviors at critical junctures during the problem-solving process. In contrast, novice problem solvers did not appear to use these metacognitive strategies and often found themselves lost in the pursuits of “wild geese.” More recently, Artzt and Armour-Thomas (in press), in their investigation of the analysis of problem solving in small groups, found that a continuous interplay of cognitive and metacognitive behaviors was necessary for successful problem solving. Other reviews of studies of mathematical problem solving (e.g., Garofalo and Lester, 1985; Silver, 1987; King and Rosenshine, 1993) suggest that a fundamental source of weakness underlying students’ performance may lie in students’ inabilities to actively monitor

78

and subsequently regulate and evaluate the cognitive processes used during problem solution. There is also some evidence about the role of cooperative learning with metacognitive scaffolding in successful mathematics performance. For example, Peterson et al. (1982, 1984) found that students’ metacognitive scaffolding for classroom learning was significantly related to their performance. Using a stimulated-recall procedure, the students were asked to recall their thoughts during mathematics instruction. They reported that they were able to judge their own understanding, to diagnose and monitor their understanding and specific cognitive processes such as reworking problems, applying information at a specific level, and checking their answers. Other researchers (e.g., Schonfeld, 1987; Xun, 2001; Kramarski et al., 2001) reported positive effects of cooperative learning with metacognitive scaffolding on students’ mathematics performance and mathematical reasoning. Kilpatrick et al. (2001) clarify that learning mathematics with understanding is not about only computation or mathematical procedures. They believe that learning mathematics with understanding has five interwoven and interdependent strands, namely Conceptual Understanding, Procedural Fluency, Strategic Competence, Adaptive Reasoning, and Productive Disposition. These five strands provide a framework for discussing the knowledge, skills, abilities, and beliefs that constitute mathematical proficiency. This framework has some similarities with the one used by Donn and Taylor (1992a) that structures different facets of mathematical or quantitative literacy (content knowledge, reasoning, appreciation of the societal impact of mathematics, and disposition), with the California Public Schools framework (1999) that includes three components (conceptsl, procedures, and

79

reasoning) of mathematical competence, and also with the one used by the National Assessment of Educational Progress, (NAEP, 2000), which features three mathematical abilities (conceptual understanding, procedural knowledge, and problem solving) and includes additional specifications for reasoning, connections, and communication. The next section discusses the mathematical proficiency strands and focus on the role of metacognitive scaffolding in teaching conceptual understanding, procedural fluency (mathematics performance), and mathematical reasoning. Since the mathematical proficiency strands are interrelated, the role of metacognitive scaffolding in teaching each strand will be briefly discussed. 2.6.1 Conceptual Understanding Conceptual understanding refers to an integrated and functional grasp of mathematical ideas. Students with conceptual understanding know more than isolated facts and methods. They understand why a mathematical idea is important and the kind of contexts in which it is useful. They have organized their knowledge into a coherent whole, which enables them to learn new ideas by connecting these ideas to what they have already learned (Donovan et al., 1999). Students with conceptual understanding are able to retrieve their knowledge and methods. That is, because they learned by connecting facts and methods under their teacher’s guidance, it is easier for them to remember and reconstruct the forgotten knowledge (Hiebert and Carpenter, 1992). When students understand a method, they are unlikely to remember it incorrectly. They monitor what they remember and try to figure out whether it makes sense. They may attempt to explain the method to themselves and correct it if necessary. When students are aware of their metacognitive thoughts, they describe their own thinking. They can realistically assess what they are capable of learning and therefore they have

80

a good sense of what they know. Also they know what they are currently doing. They have self-regulation, they are keeping track of what they are doing and knowing well how to use their previous knowledge to guide the problem solving actions. So when students are metacognitively trained, they are likely to learn with conceptual understanding. A significant indicator of conceptual understanding is being able to represent mathematical situations in different ways and knowing how different representations can be useful for different purposes. To find one’s way around the mathematical terrain, it is important to see how the various representations connect with each other, how they are similar, and how they are different. The degree of student’s conceptual understanding is related to the richness and extent of the connections they have made (Kilpatrick et al, 2001). Students who are working cooperatively and scaffolded metacognitively ask themselves about the construction of relationships between the previous and the new knowledge (What are the similarities / differences between the problem at hand and the problems solved in the past?) (Kramarskis et al., 2001), therefore they can identify the similarities and differences between the various strategies used. So doing might help students to represent the problem in various representations, and by connecting these representations with each other, they are likely to gain the conceptual understanding and then to construct the correct solution. Schonfeld (1987) and King and Rosenshine (1993) found that when students were metacognitively trained they could make connections between mathematical concepts in different areas. Conceptual understanding helps students avoid many critical errors in solving problems, particularly errors of magnitude. For example, “if they are multiplying 9.83

81

and 7.65 and get 7519.95 for the answer, they can immediately decide that it cannot be right. They know that 10 x 8 = 80, so multiplying two numbers less than 10 and 8 must give a product less than 80. They might then suspect that the decimal point is incorrectly placed and check that possibility” (Kilpatrick et al., 2001, p.6). Students who work cooperatively and are scaffolded metacognitively are likely to understand the concept of addition, connect the current problem with the previous one, represent the problem in different ways, expect the product, and check their learning strategy and the product, and therefore, they are unlikely to do critical errors. 2.6.2 Procedural Fluency Procedural fluency refers to knowledge of procedures or algorithms, knowledge of when and how to use them appropriately, and skill in performing them flexibly, accurately, and efficiently (Kilpatrick et al, 2001). Flexibility requires the knowledge of more than one approach to solving a particular kind of problem, such as two-digit multiplication. Students need to be flexible in order to choose an appropriate strategy for the problem at hand, and also to use one method to solve a problem and another method to double-check the results. Accuracy depends on several aspects of the problem-solving process, among them careful recording, knowledge of number facts and other important number relationships, and double-checking results. Efficiency implies that the student does not get bogged down in too many steps or lose track of the logic of the strategy. An efficient strategy is one that the student can carry out easily, keeping track of sub problems and making use of intermediate results to solve the problem (Russell, 2000). Students within cooperative learning with metacognitive scaffolding setting are likely to represent the problem in different ways where they can select the appropriate

82

approach or procedure to solve the problem. That is, they ask themselves about the appropriate approach to solve the problem (What are the appropriate approach / strategy / procedures to solve the problem? (Kramarskis et al., 2001). They plan their learning by understanding the whole problem before getting start, keep track of what has been done, comparing the differences and similarities of the current problem and the problems have been solved, and evaluate the outcome of any action. Therefore these students are unlikely to make mistakes during procedures application. They practice problems in a way that requires different types of procedures (Mathematics Framework for California Public Schools, 1999). For example, they mix subtraction and addition operations. This type of practice provides students with an opportunity to understand better how different procedures work by making them think about which is the most appropriate procedure for solving each problem. In other words, they represent different procedures and evaluate the outcomes then select the appropriate one and justify their selection. Research indicates that long-term retention of mathematics procedures requires frequent refreshers at different points in the students’ mathematical learning (Bahrick et al., 1993). Students who work cooperatively with metacognitive scaffolding always ask themselves questions like: what the whole problem is about, what are the similarities and differences between the current problem and the problems already were solved, and what is the appropriate procedure to solve the current problem. Therefore, they are likely to refresh their mathematical learning and procedures. Moreover, metacognitively trained students can modify or adapt procedures to make them easier to use (Carpenter et al., 1998). For example, students who work individually with limited metacognitive strategies in fractions addition, would

83

ordinarily need paper and pencil to add

2 1 and , while students within cooperative 4 2

learning with metacognitive scaffolding setting recognize that

2 1 equals to and 4 2

therefore they add

1 1 to mentally to get 1 as the result. 2 2

2.6.3 Strategic Competence Strategic competence refers to the ability to formulate mathematical problems, represent them, and solve them. So for students to solve mathematical problems they need to formulate the problem first then they can use mathematics to solve it. In other words they need experience and practice in both problem formulating and problem solving. Therwefore, they should know a variety of solution strategies as well as determining which strategies might be useful for solving a specific problem. Students with experience in solving mathematical problems and with limited or without formulating experience usually encounter difficulties in figuring out exactly what the problem is (Kilpatrick et al, 2001). Students within the cooperative learning with metacognitive scaffolding setting are encouraged to understand the whole problem first. They are encouraged to ask themselves what the whole problem is about, represent the problem in different ways and connect these representations, determine the similarities and differences between

84

the problem on hand and others they have solved, select the appropriate approach to solve the problem, and evaluate the outcomes (Kramarskis et al., 2001). That is, they are encouraged to plan (before the solution), monitor (during the solution), and to evaluate (after the solution). So doing assisted them to formulate, represent, and solve the problem. For example, if they encounter the following purchase problem: 3 1 Dinar. Ali’s shop price is Dinar 4 3

“At Ahmad’s shop the price of a piece of cake is

less than Ahmad’s price. How much 3 pieces of cake cost at Ali’s shop?” They may identify what the problem is about by studying the relationships among the variables in the problem and determine what is known and what to be found. By doing so, they are likely to conclude that subtraction and multiplication initially should be used (problem formulation). With a formulated problem in hand, students within the cooperative learning with metacognitive scaffolding setting are more likely to represent it mathematically in some fashion, whether numerically, symbolically, or graphically. They build a mental image of the problem’s essential components. They avoid selecting numbers and preparing to perform arithmetic operations on them directly. Rather they are likely to construct a mental model of the variables and relations described in the problem. That is, they generate a mathematical representation of the problem that captures the core mathematical elements and ignore the irrelevant features (Kilpatrick et al, 2001). For the purchase problem, students within the cooperative learning with metacognitive scaffolding setting may draw a number line and locate each cost per piece of cake on it to solve the problem. They may represent the problem by transforming the two

85

fractions into equivalent fractions with a like denominator ( problem

9 4 , ) to solve the 12 12

(problem presentation). Having many different representations students try

to connect the relationships among them by determining the common mathematical structures. That is, they focus on structural relationships that provide the clues for how the problem might be solved (Hagarty et al., 1995). They compare the current problem with the previous one. For example, they recognize that this problem relates to subtraction two fractions with unlike denominators and this is different from what they solved previously with like denominator fractions. Therefore they are likely to conclude that they cannot subtract the numerators directly and try to find equivalent fractions with like denominators. Students with strategic competence need to choose flexibly among the proposed approaches to suit the demands presented by the problem and the situation in which it was posed. Flexibility of approach can be seen when a method is adjusted or created to fit the requirement of the problem (Kilpatrick et al., 2001). Students within the cooperative learning with metacognitive scaffolding setting are flexible in their approaches (Butler, 1995). They are likely to represent the problem in different forms and therefore they have different solution approaches, and by comparing these approaches they can select the appropriate one that will be evaluated to check its appropriateness. For the purchase problem, they may select the approach of transforming the two fractions into equivalent fractions with a like denominator as the

appropriate approach to solve the problem. They may offer fractions with a common denominator and formulate:

9 4 and as equivalent 12 12

86

9 4 5 = as the difference in price for a piece of cake. For 3 pieces of cake, the 12 12 12

formulation would be:

5 5 5 15 + + = 12 12 12 12 = 5 4 1 4

=1

Because students within the cooperative learning with metacognitive scaffolding setting are encouraged to ask evaluation questions, they are likely to check if the

solution makes sense by testing if

9 3 4 1 equals to and equals to . They may 12 4 12 3

draw a rectangle and divide it into 4 equal pieces and shade 3 pieces, and then they divide the rectangle into 12 equal pieces where they find that the three shaded pieces make nine pieces and so on for the other fractions (problem solving and evaluation). Sternberg (1986b) refers to metacognitive strategies as Metacomponents.

Metacomponents are executive processes that control other cognitive components as well as receive feedback from these components. According to Stemberg (1986b), Metacomponents are responsible for “figuring out how to do a particular task or set of tasks, and then making sure that the task or set of tasks are done correctly” (p. 24). These executive processes involve planning, monitoring and evaluating problemsolving activities. Research indicates that metacognitively scaffolded students are more strategic and perform better than untrained students (Garner and Alexander, 1989). One explanation is that metacognitive scaffolding allows individuals to plan, sequence, monitor, and evaluate their learning in a way that directly improves performance (Schraw and Dennison, 1994).

87

2.6.4 Adaptive Reasoning Kilpatrick et al. (2001) refers to adaptive reasoning as the students’ ability to think logically about the relationships among the mathematical concepts and situations. This reasoning stems from careful consideration of alternatives, and includes knowledge of how to justify the conclusions. Adaptive reasoning is the glue that holds everything together, the lodestar that guides learning. Students use it to navigate through the many facts, procedures, concepts, and solution methods and to see that they all fit together in some way, that they make sense. Adaptive reasoning is much broader than formal proof and other forms of deductive reasoning, it includes not only informal explanation and justification, but also intuitive and inductive reasoning based in pattern, analogy, and metaphor. Since learning according to Piaget occurs by assimilation and/or accommodation through resolving the cognitive conflicts (Mugny and Doise, 1978), in learning mathematics, reasoning is used to settle disputes and disagreements and then knowledge is changed or reconstructed. Mathematical answers are right because they follow from some agreed upon assumptions through series of logical steps (Kilpatrick et al., 2001). Mathematics Framework for California Public Schools (1999) identifies the mathematical reasoning steps as follows: a) making decisions about how to approach problems through analyzing the problem by identifying relationships, distinguishing relevant from irrelevant information, sequencing and prioritizing information, and observing patterns, and determine when and how to break the problem into simpler parts. b) using strategies, skills, and concepts in finding solutions through using

88

estimation to verify the reasonableness of calculated results, applying strategies and results from simpler problems to more complex problems, using a variety of methods, such as words, numbers, symbols, charts, graphs, tables, diagrams, and models, to explain mathematical reasoning, expressing the solution clearly and logically by using the appropriate mathematical notation, terms and clear language; support solutions with evidence in both verbal and symbolic work, indicating the relative advantages of exact and approximate solutions to problems and give answers to specified degree of accuracy, and making precise calculations and check the validity of the results from the context of the problem. c) moving beyond a particular problem by generalizing to other situations through evaluating the reasonableness of the solution in the context of the original situation, realizing the methods of deriving the solution and demonstrate a conceptual understanding of the derivation by solving similar problems, and developing generalizations of the results obtained and applying them in other situations. Pollack (1997) indicates that mathematical reasoning plays a significant role in student’s ability to take an open-ended question and transform it into unambiguous something to solve, that is, to formulate and summarize the real-life problem. In summary, mathematical reasoning refers to the students’ ability to identify the similarities and differences among facts, concepts, procedures, and situations and then think logically about the relationships among them. After the relationships were identified, appropriate strategies for solving the problem are selected and reasonable reasons about strategies selection and calculated results are provided. Finally, justified strategies and results are applied in other situations (i.e., strategies generalization and solving real-life problems).

89

It is apparent from the steps of mathematical reasoning identified by the mathematics framework for California public schools that mathematical reasoning comprises both strategic competence and adaptive reasoning of Kilpatrick’s et al. (2001) model of mathematical proficiency. Thus, the mathematical reasoning term in this study was used to indicate strategic competence and adaptive reasoning simultaneously. 2.6.5 Productive Disposition Resnick (1987) refers to the productive disposition as the tendency to see mathematics as sensible, useful, and worthwhile, coupled with a belief in diligence and one’s own efficacy. For students to learn mathematics with understanding (proficiently), they have to believe that mathematics is understandable, not arbitrary, that, with diligent effort, it can be learned and used, and that they are capable of figuring it out. It is counterproductive for students to believe that there is some mysterious factor that determines their success in mathematics (Kilpatrick et al, 2001). Therefore, learning mathematics with understanding goes beyond being able to understand, compute, and solve problems. It takes account of a disposition toward mathematics that is personal. Students’ disposition toward mathematics may play a crucial role in their understanding and success. For instance, Dweck (1986) indicates that students who view their mathematical ability as fixed and test questions as measuring their ability rather than providing opportunities to learn are likely to avoid challenging tasks and be easily discouraged by failure, whereas students who view ability as expandable in response to experience and training are more likely to seek out challenging situations and learn from them.

90

Kilpatrick et al. (2001) attribute the development of productive disposition to the development of the other mathematical proficiency strands. For example, when students build their own strategies to solve the mathematical task, their attitudes and beliefs about themselves as mathematics students become more positive. In addition, the more mathematical concepts they understand, the more sensible mathematics becomes. In contrast, when students are seldom given challenging mathematical tasks to do, they come to expect that memorizing rather than sense is the appropriate approach to learn mathematics, and they begin to lose confidence in themselves as students. Similarly, when students see themselves as capable to operate the mathematical procedures and reason mathematically, their disposition is more likely to be positive. Since students within the cooperative learning with metacognitive scaffolding setting are more likely to have conceptual understanding, procedural fluency, strategic competence, and adaptive reasoning, they seem to have a positive attitudes and beliefs. Also, when students apply the metacognitive strategies within the cooperative learning environment, they discuss, share, and contrast their ideas and their teacher ideas. This conflict environment is likely to enhance students to see themselves as capable to learn with understanding (Cobb et al., 1995), which in turn, seems to help them to have a positive attitudes and beliefs.

2.7 Cooperative Learning with Metacognitive Scaffolding and Mathematical Reasoning Students within the cooperative learning with metacognitive scaffolding setting are more than likely to reason mathematically in their learning situations. They are guided about the knowledge of when, where, and why to use the strategies for problem

91

solving (Pressley and McCormick, 1987). As metacognitive scaffolding comprises planning, monitoring, and evaluation, metacognitively trained students are likely to plan, monitor, and evaluate their learning strategies and solutions. Planning is essential to formulate, identify, and define the problem and then building the relationships among its concepts and procedures. To select the appropriate strategies, the students need to regulate or monitor their problem performance by self-generating feedback. Evaluation enhances students to reflect on their solutions or alternatives so as to direct their future steps (Jacobs and Paris, 1987). Since mathematical reasoning requires thinking about the relationships between mathematical facts, concepts, and situations, students within the cooperative learning with metacognitive scaffolding setting are enhanced to identify the similarities and differences between the current problem and the ones they have already solved. Doing so is likely to enable them to compare concepts, procedures, and strategies which in turn, enable them to establish the relationships among them. For example, when the

students encounter the following task

3 2 + to solve, the teacher asks: what are the 4 5

differences / similarities between the current task and those you solved last class? What in your prior knowledge will help you in this particular task? What you should think about first? By answering these questions, students may possibly reach to the conclusion that the current task is regarding adding two fractions with unlike denominators. Studies conducted by Chi et al. (1994); Mevarech and Kramarski, (in press); Slavin (1996); Cossey’s (1997); and Webb (1989) showed that metacognitive scaffolding is one of the best means for making connections between mathematical facts, concepts, and procedures.

92

While mathematical reasoning requires selecting appropriate strategies for solving the task and justifying both the strategies’ selection and the task’s solution, students within the cooperative learning with metacognitive scaffolding setting seem to be able to select and justify the appropriate strategies for solving the task. For the adding fractions with unlike denominator example, the teacher asks metacognitive questions that help students to plan, monitor, and evaluate their learning such as: Is the sum would be less than or greater than 1? Thinking to answer this question is likely leading students to answer “the sum will be greater than 1”, and then the teacher asks why?

By relating to their previous knowledge, students may justify that

2 1 is greater than , 5 4

and

3 1 + equals to 1, so the sum will be greater than 1. The teacher then asks what 4 4

are the appropriate strategies to find the sum of these two fractions? What should you do first? Students with metacognitive scaffolding are likely to respond that they can not add directly unless they make the two denominators equal. How you should do so? The teacher asks. The students are likely to compare this task with the previous ones

and relate it to the equivalent fractions and offer

15 8 and as equivalent fractions 20 20

with a common denominator. The teacher asks students to justify why they have chosen 20 as the common denominator. By relating to the Least Common Multiple (LCM), students seem to respond that 20 is the smallest number that both five and four go into. How did you come up with that? the teacher asks. Again through understanding and relating to the LCM, students are likely to answer “by multiplying 5 and 4. The teacher then asks students to write down the processes of solving the

93

task. Within the cooperative learning with metacognitive scaffolding setting, students are likely to write the solution as follows: 3 2 + 4 5 3×5 2×4 3×5 + 2 × 4 + = 4×5 5× 4 4×5 = 15 + 8 20 23 20

=

=

The teacher asks, are you in the right track? How do you check if the solution makes sense? How do you know that you have added the same fractions as in the original task? How well did you do? The students may use different representations (graphs,

models, symbols, numbers line, etc) to prove that

3 15 2 equivalent to and 4 20 5

equivalent to

8 . Hoek et al. (1999) and Mevarech (1999) studies showed that 20

metacognitive scaffolding is effective for developing the selection of the appropriate

strategies for solving the problem. Finally, mathematical reasoning requires applying strategies in other situations. Students within the cooperative learning with metacognitive scaffolding setting are more likely to generalize their learning strategies to other situations. Based on the adding two fractions with unlike denominators example, the teacher asks: how might you apply this line of thinking to other situations? Could you derive a rule that would work for adding or subtracting any fractions with unlike denominators? The teacher then provides different tasks, word-problems, and real-life problems regarding adding

94

and subtracting fractions with unlike denominators and asks the students to solve them through applying the same line of thinking they applied before.

2.8 Cooperative Learning with Metacognitive Scaffolding and Real-Life Problem Solving Real-life problems are problems that people encounter in everyday life. They are generally problems in which one or several aspects of the situation are not specified. The general nature of these problems is that the goals are vaguely defined or unclear (Voss and Post, 1988), their descriptions are not clear, and the information needed to solve them is not entirely contained in the problem statements; consequently, it is not obvious what actions to take in order to solve them (Chi and Glaser, 1985). Real-life problems entail multiple solutions, solution paths, or no solutions at all (Kitchner, 1983). Since real-life problem solving may generate a large number of possible goals, Sinnott (1989) insists that the solvers must have a mechanism or strategies for selecting the best goal or solution. Hong (1998) summarized the processes of real-life problems which their goals are vaguely defined into three processes: (a) representation problem, (b) solution processes, and (c) monitoring and evaluation. A representation problem is established by constructing a problem space, including defining problems, searching and selecting information, and developing justification for the selection. The solution process involves generating and selecting solutions. Finally, the monitoring and evaluating process requires assessing the solution by developing justifications for it. Since reallife problems usually have no clear goals and require the consideration of alternative solutions as well as competing goals, solving this kind of problems requires students to regulate the selection and execution of a solution process. That is, when goals or

95

action alternatives are unclearly defined, students have to organize and direct their cognitive endeavors in different ways. Because students need and use mathematics in their everyday lives, it is very critical

to learn solving real-life problems. For instance, students learn the concept of

1 2

effectively by solving a problem like if two ice cream cones cost 10 cents, how much does one cone cost? That is, learning to solve real-life problems produces active learning and easily retrieved knowledge as Brown et al. (1989) assert that when learning includes real-life problems, students acquire content and skills through the resolution of problems. It is not that traditional teaching practices do not use examples, real-life problems and other devices. It is that the overall approach is turned around the wrong way. Students are taught the isolated basics and then are expected to apply them to artificial problems (Tiene and Ingram, 2001). Lesh (1985) indicates that getting a collection of isolated concepts in a student’s head (e.g., measurement, addition, multiplication, decimals, proportional reasoning, fractions, negative numbers) does not guarantee that these ideas will be organized and related to one another in some useful way; it does not guarantee that situations will be recognized in which the ideas are useful or that they will be retrievable when they are needed. Tiene and Ingram (2001) assert that the best approach of teaching is to ground all learning as much as possible in tasks, activities, and problems that are meaningful to the students. If it is important for students to learn facts, they will learn them most effectively while engaged in meaningful tasks.

96

For learning mathematics, when situations are mathematized in the classroom such as balancing a class budget, the students will be engaged in multiple mathematics processes and they will learn how mathematical concepts are related to one another in a useful and meaningful way. Such experiences also require students to talk and think about mathematics with one another and with the teacher (Lesh, 1985). Within the cooperative learning with metacognitive scaffolding setting, students are enhanced to solve real-life problems because they are encouraged too plan, to monitor problem-solving processes, to reflect on the goals and solution processes and to construct cogent arguments for their proposed solutions. In a study of history experts, Wineburg (1998) found that planning, monitoring, and evaluation helped students to solve a real-life problem in the absence of domain knowledge. Lin and Lehman (1999); Davis and Linn (2000); King (1991a, 1991b); Palincsar and Brown (1984, 1989; and Kramarski et al., 2002) found that planning, monitoring, and evaluation enhanced metacognitive knowledge and reflective thinking which enhanced the processes of solving real-life problems. Students within the cooperative learning with metacognitive scaffolding setting are likely to reason and defend their selections and solutions. As students select a good solution from among the many viable solutions, they provide the most viable, the most defensible and the most cogent argument to support their preferred solution, and defend it against alternative solutions (Jonassen, 1997; Voss and Post, 1988). In addition, students within the cooperative learning with metacognitive scaffolding setting may also evaluate their selection by examining and comparing other alternatives. Sinnott (1989) noted that during the process of solving a real-life problem, successful students planed, monitored, and evaluated their own processes

97

and movements from state to state, as well as select information, solutions, and emotional reactions.

2.9 Cooperative Learning with Metacognitive Scaffolding and Motivation Motivation is the reason why an individual behaves in a given manner in a given situation. It exists as part of one’s goal structures, one’s beliefs about what is important, and it determines whether or not one will engage in a given pursuit (Ames, 1992). Two distinct types of academic motivation interrelate in most academic settings, intrinsic and extrinsic motivation. Academic intrinsic motivation is the drive or desire of the student to engage in learning “for its own sake.” Students who are intrinsically motivated engage in academic tasks because they enjoy them. They feel that learning is important with respect to their self-images, and they seek out learning activities for the sheer joy of learning (Middleton, 1992, 1993a). Their motivations tend to focus on learning processes such as understanding and mastery of mathematical concepts (Duda and Nicholls, 1992). When students engage in tasks in which they are motivated intrinsically, they tend to exhibit a number of pedagogically desirable behaviors including increased time on task, persistence in the face of failure, more elaborative processing and monitoring of comprehension, selection of more difficult tasks, greater creativity and risk taking, selection of deeper and more efficient performance and learning strategies, and choice of an activity in the absence of an extrinsic reward (Lepper, 1988). On the other hand students who are extrinsically motivated engage in academic tasks to obtain rewards (e.g., good grades, approval) or to avoid punishment (e.g., bad grades, disapproval). These students’ motivations tend to focus on learning products as

98

obtaining favorable judgments of their performance from teachers, parents, and peers or avoiding negative judgments of their performance (Ames, 1992). Teachers often complain that their students are not motivated and hence cannot or do not learn well (Driscoll, 1998). The motivation that teachers wish their students to have is intrinsic motivation. Intrinsic motivation is important because it contributes to learning processes and achievement, but it is also important as an outcome (Ames, 1990). She adds “Effective teachers are those who develop goals, beliefs, and attitudes in students that will sustain a long-term involvement and that will contribute to quality involvement in learning” (p.413). There are many ways to promote motivation. Garrison (1997) expounds that “to direct and sustain motivation students must become active students” (p. 8). Task motivation is integrally connected to task control and self-management. Driscoll (1998) declares “Motivation appears to be enhanced when students’ expectancies are satisfied, when they attribute their successes to their own efforts and effective learning strategies, and when the social climate fosters interaction and cooperation among students” (p. 312). According to constructivist point of view, learning is an active process in which the students construct their own knowledge. So they encounter difficult problems that they cannot solve by using only their current knowledge. In this case, students are challenged by the task and they will be motivated more. Bruner (1973) maintain that student may be motivated more quickly when given a problem they cannot solve, than they are when given some little things to learn on the promise that if they learn these well, three weeks later they will be able to solve an exciting problem (Shulman, 1973). Hein (1991) indicates that motivation is a key component in learning. Not only is it the case that motivation helps learning, it is essential for learning. Motivation is

99

broadly conceived to include an understanding of ways in which the knowledge can be used. Unless the student knows the reasons why, he or she may not be very involved in using the knowledge that may be instilled in him or her. Even by the most severe and direct teaching. Motivation is particularly important in learning mathematics with understanding. When students are mathematically motivated, they will perceive mathematics as useful and worthwhile and then they will see themselves as effective students of mathematics (Resnick, 1987). Students who view their mathematical ability as expandable in response to experience and training are more likely to seek out challenging situations and learn from them. In contrast, students who view their mathematical ability is fixed are likely to avoid challenging problems and be easily discouraged by failure (Dweck, 1986). Mathematics instruction plays an important role in encouraging students’ motivation to learn mathematics. Students within cooperative learning with metacognitive scaffolding setting, are likely to be motivated more to learn mathematics. These students are encourager to negotiate among themselves the norms of conduct in the class, and when those norms allow students to be comfortable in doing mathematics and sharing their ideas with others, they see themselves as capable of understanding and then doing mathematics effectively (Cobb and Yackel, 1995).

CHAPTER THREE

METHODOLOGY

100

3.1 Introduction The purpose of this study was to investigate the effects of cooperative learning with metacognitive scaffolding (CLMS) and cooperative learning (CL) methods on mathematics performance, mathematical reasoning, and metacognitive knowledge among high and low ability fifth-grade students in Jordan. This chapter discusses the methodology that was used in this study. It describes in detail the population and sample, the experimental conditions, the research design, the instructional materials and instruments, the procedures, and data analysis procedure and the method that was used in the analysis of data. It is important to note that everyday classroom instructions and all reading materials (except for the English subject) used in the participating schools are in the Arabic Language. Therefore, all the materials and instruments used in this study were translated into Arabic.

3.2 Population and Sample The population of this study comprised male fifth grade students enrolled in the first public educational directorate in Irbid Governorate in the first semester for the academic year 2002 / 2003. The first public educational directorate in Irbid Governorate includes 44 male primary schools. Public schools in Jordan are not coeducational. In order to implement this study in a naturalistic school setting, existing intact classes were used (O’deh and Malkawi, 1992). The sample consisted of 240 male students who studied in six fifth-grade classrooms and were randomly (simple random sample)

101

selected from three different male primary schools i.e., two classes from each school. The three schools were also randomly selected from the primary schools where mathematics was taught in heterogeneous classrooms with no grouping or ability tracking. The size of the classes was approximately similar, and the mean age of the students was 10.6 years. Students in the selected schools – as well as all Irbid Government schools - were from approximately equivalent socioeconomic status as defined by the Jordan Ministry of Education. Each of the three male teachers who participated in this study taught two classrooms. All the teachers were men who had similar levels of education (B.Ed. major in mathematics), had more than 7 years of experience in teaching mathematics, and had taught in heterogeneous classrooms. The teachers who taught the experimental groups were exposed to one week training on the instructional methods. The participating students were informed that the purpose of this study was to examine different learning strategies that may help in the improvement of students’ mathematics performance, mathematical reasoning, and metacognitive knowledge.

3.3 Experimental Conditions The three schools were assigned randomly to one of the following conditions: 1. CLMS: students taught mathematics via the Cooperative Learning with Metacognitive Scaffolding method (n = 80). 2. CL: Students taught mathematics via Cooperative Learning with no Metacognitive Scaffolding (n = 79). 3. T (control group): students taught mathematics via the present classroom practice (traditional method), that is, without Metacognitive Scaffolding or Cooperative Learning methods (n = 81). (See table 3.1).

102

**Table 3.1: Mechanisms for the three groups
**

Group 1 (CLMS) N = 80 Group 2 (CL) N = 79 Group 3 (T) N = 81 The whole class with neither cooperative learning nor Metacognitive Scaffolding

Cooperative Learning with metacognitive scaffolding

Cooperative Learning with no Metacognitive Scaffolding

Cooperative learning Students worked, discussed, interacted in groups, and asked themselves and their members metacognitive questions

Cooperative Learning Students worked, discussed, and interacted in groups with no MQ

Without Cooperative Learning

Metacognitive Scaffolding a) The teacher asked metacognitive questions{MQ} and coached students to ask MQ b) Students used metacognitive questions cards

Without teacher’s metacognitive scaffolding Without metacognitive questions cards

Without teacher’s metacognitive scaffolding Without metacognitive questions cards

The three groups were different from one another in terms of the instructional method and materials used. The CLMS group was asked metacognitive questions by the teacher and students in this group used metacognitive questions cards in cooperative learning setting. The CL group studied cooperatively with neither teacher’s metacognitive questions nor using metacognitive questions cards, whereas the T group studied in the usual manner with neither cooperative learning, teacher’s metacognitive questions, nor metacognitive questions cards.

3.4 Research Design

103

This quasi-experimental study was designed to investigate the effects of cooperative learning with metacognitive scaffolding and cooperative learning methods on mathematics performance, mathematical reasoning, and metacognitive knowledge. The study employed Factorial Design 3x2. It was designed to investigate the effects of the independent variable on the dependent variables at each of the two levels of the moderator variable. The research design is illustrated in table 3.2. Table 3.2: Research Design

Moderator Variable (Ability) High-ability (Y1) Low- ability (Y2) Independent Variable (Instructional Method) CLMS 1 4 CL 2 5 T 3 6

O1 X1 O3 X2 O5 X0 O7 X1 O9 X2 O11 X0

Y1 Y1 Y1 Y2 Y2 Y2

O2 O4 O6 O8 O10 O12

(1) (2) (3) (4) (5) (6)

=

X1: CLMS X2: CL X0: T Y1: High-ability Y2: Low-ability

O1 = O3 = O5 = O7 O2 = O4 = O6 = O8

O9 = O11 = Pre-test.

= O10 = O12 = Post-test.

The independent variable of this study was the instructional method with three categories: 1. Cooperative learning with metacognitive scaffolding instructional method (CLMS). 2. Cooperative learning instructional method (CL). 3. Traditional instructional method (T).

104

The moderator variable was the ability level with two categories: 1. High-ability. 2. Low-ability. The dependent variables were: 1. Mathematics performance (MP). 2. Mathematical reasoning (MR), and 3. Metacognitive knowledge (MK). The design of the present study compares three instructional methods (a) cooperative learning with metacognitive scaffolding instructional method, (b) cooperative learning with no metacognitive scaffolding instructional method, and (c) traditional instructional method with neither cooperative learning nor metacognitive scaffolding. Slavin (1996) recommended the use of such research design because it enables researchers to hold constant all factors other than the ones being studied.

3.5 Instructional Materials and Instruments 3.5.1 Instructional materials In order to study the students’ mathematics performance, mathematical reasoning, and metacognitive knowledge in a naturalistic setting of the classroom, the instructional

105

materials used in this study were based on the fourth unit from the mathematics textbook (Adding and Subtracting Fractions) designed by the Ministry of Education for all fifth-grade students in Jordan, teacher’s lesson plans, and metacognitive questions card. 3.5.1.1 Adding and Subtracting Fractions Unit “Adding and Subtracting Fractions” was the unit chosen for this study. This particular unit was chosen for two reasons: a) Jordan eighth-students’ performance regarding fractions according to the TIMSS-R (1999) findings was very low and students committed many undetermined errors. This low performance in fractions, particularly addition and subtraction of fractions demands Jordanian educators to pay attention to this particular topic. The TIMSS-R study was conducted on eighth-grade students but according to the Jordanian curriculum the topic of fractions is taught in the fifth-grade. Jordanian students start learning the basics of adding and subtracting fractions in the fifth-grade; and b) This topic was scheduled by the schools to be covered by the teacher in early December 2002 / 2003, which is also the same duration of time planned for this study. The “Adding and subtracting Fractions” unit consists of 12 lessons, which are, Introduction to Fractions, Mixed Numbers, Equivalent Fractions, Simplifying Fractions, Comparing and Ordering Fractions and Mixed Numbers, Adding Fractions, Adding Mixed Numbers, Adding Fractions Problem Solving, Subtracting two

Fractions which one Fraction’s Denominator is Multiple of the second Fraction’s Denominator, Subtracting two Fractions which one Fraction’s Denominator is not Multiple of the second Fraction’s Denominator, Subtracting Mixed Numbers, and Subtracting Fractions Problem Solving respectively. Within each school, the teacher

106

conducted the class according to his assigned teaching method for 15 sessions. Each lesson started with the new topic explanation and followed by mathematical exercises and problem solving. One session (45 minutes) was conducted to teach the first ten lessons and two sessions to teach the last two lessons as planned from the Ministry of Education in teacher’s guide. All sessions were presented from written lesson plans to ensure that all participating students in the three groups received the same quantity of knowledge. 3.5.1.2 The Metacognitive Questions Cards A set of metacognitive questions cards (see Appendix A) was developed by the researcher based on the metacognition components (planning, monitoring, and evaluation) designed by Jacobs and Paris (1987); and North Central Regional Educational Laboratory, (NCREL, 1995). The students taught via the CLMS method used the questions cards to scaffold their learning processes when they engaged in cooperative learning activities with their respective peers. These questions were categorized into the following groups of metacognitive questions: Planning: “What is the problem all about?” “What are the strategies we can use to solve the problem and why?” (There were 8 questions in this category). Monitoring: “Are we on the right track?” (There were 9 questions in this category). Evaluation: “What explanations can we make and what evidence do we have to justify that our solution is the most viable?” (There were 5 questions in this category). These questions were closely paralleled to the Kilpatrick’s model of mathematical proficiency (2001). They consisted of questions supporting mathematical proficiency and mathematical reasoning skills, such as “what”, “how”, and “why” as well as

107

questions that were found in King’s generic question stems (King, 1991b). These questions were designed to facilitate students’ understanding of domain knowledge, to develop metacognitive thinking, and to develop mathematical reasoning, such as questions regarding making decisions about approaching the problem, selecting the appropriate strategies to solve the problem, and regarding generalizing the solution processes to other situations. The students taught via the CLMS method were instructed

and reminded frequently to think about the questions, and use the questions to facilitate their problem solutions.

3.5.2 Instruments In this study, two major instruments were used to assess students’ mathematics performance, mathematical reasoning, and metacognitive knowledge. A mathematics achievement test was used to assess students’ mathematics performance and mathematical reasoning and a metacognitive questionnaire was used to assess students’ metacognitive knowledge. 3.5.2.1 The Mathematics Achievement Test (The pre-Test and post-Test) The mathematics achievement test administered by the three groups’ participants in this study was adapted from the mathematical competency test developed by Jbeili (1999). The test-retest reliability coefficient of that test was .93. The test-retest approach of measuring reliability is considered the best approach that provides the test’s consistency over time (Tuckman, 1999). The mathematical competency test consisted of 5 conceptual understanding items, 11 procedural fluency items, and one problem solving regarding adding and subtracting fractions. Since there were no mathematical reasoning items included in that test, the researcher constructed these respective items (see appendix B) based on Kilpatrick’s model (2001), NCTM

108

standards (2000), and Mathematics Framework for California Public Schools (1999). With these new added items, the reliability coefficient was measured by applying the test on a pilot sample. The pre-test and post-test questions were similar in content but their order and numbering were randomized. The mathematics achievement test questions consisted of 24 mathematical items and sub items and a real-life problem. The mathematics achievement test questions covered the following topics: equivalent fractions, simplifying fractions, comparing and ordering fractions and mixed numbers, adding and subtracting fractions, and adding and subtracting mixed numbers. Three constructs, which tightly correspond to the mathematics performance and mathematical reasoning, were identified as important for measuring mathematics performance and mathematical reasoning for this study: (a) conceptual understanding, (b) procedural fluency, and (c) mathematical reasoning. The mathematics achievement test questions were composed of four kinds of items. One kind (10 items) was based on multiple-choice items regarding conceptual understanding and procedural fluency. The second kind (6 items) was based on open-ended tasks regarding conceptual understanding and procedural fluency. The third kind (8 items) was specifically designed to assess students’ mathematical reasoning. The mathematical reasoning items were designed to require students to go beyond presenting facts to thinking about those facts. The 8 items asked students to estimate the results, explain the solution clearly, and justify and support the solutions with evidence. The fourth kind was a real-life problem that involved conceptual understanding, procedural fluency, and mathematical reasoning. The problem asked students to decide the better buy from two different prices and quality of mixed fruit juice. The student had to calculate the mixed fruit juice volume in each shop, compare the prices and quality, decide the

109

better buy, and provide reasons for his decision. To gain a deeper understanding of students’ mathematical reasoning, the items regarding mathematical reasoning and the mathematical reasoning criteria in the real-life problem were separately analyzed following the method used by Kramarski et al. (2001). 3.5.2.2 The Scoring of Mathematics Achievement Test The total score of the test was 44. The distribution of the mathematics performance and mathematical reasoning items and their scores across conceptual understanding (CU), procedural fluency (PF), and mathematical reasoning (MR) is illustrated in appendix C. The 24 mathematics items and sub items and the real-life problem scoring were as follows: Multiple-choice items: For each item, students received a score of either 1 (correct answer) or 0 (incorrect answer), and a total score ranging from 0 to 10. Open-ended task items: For each item, students received a score of either 1 (correct answer) or 0 (incorrect answer), and a total score ranging from 0 to 6. Mathematical reasoning items: The scoring procedure is adopted from

Kramarski et al. (2001) and has a repeated .90 interjudge reliability. For each item, students received a score between 0 and 2, and a total score ranging from 0 to 16. For

example, “In the following item,

9 5

…

2 , explain which sign >, <, or = that will 3

make the statement true.” A score of 0 indicates incorrect selection and explanations

or explanations that are irrelevant to the task (e.g.,

9 5

<

2 because when the 3

110

denominator is smaller the value is greater. Nothing is mentioned about numerator or transferring it into a common denominator). A score of 1 indicates an explanation that has some satisfactory elements but may has omitted a significant part of the task (e.g.,

9 2 > because when transformed into a common denominator the numerator 27 is 5 3 bigger than the numerator 10. Nothing is mentioned about the denominators. A score of 2 indicates a clear, unambiguous explanation of student’s mathematical reasoning

(e.g.,

27 9 2 > , when transform into equivalent fractions with a like denominator 5 3 15

and

10 , the fraction with the larger numerator is the larger fraction if the denominators 15

are the same, since the denominators (15, 15) are same and the numerator 27 is bigger

than the numerator 10,

27 10 > . 15 15

The real-life problem: A scoring rubrics (see appendix D) was adapted from the Kramarski et al. (2001) procedure with a repeated .86 interjudge reliability. Four criteria, which tightly correspond to the conceptual understanding, procedural fluency, and mathematical reasoning, were identified as important for measuring students’ ability to solve the real-life problem. Students’ answers were scored on these criteria, each criterion ranges from 0 (no solution) to 3 (highest level solution), and a total score ranging from 0 to 12. The criteria were:

111

1. Referencing all data (referring to all data in each of the two offers: mixed fruit juice volume, components, and prices. Identifying relationships, distinguishing relevant from irrelevant information- Mathematical Reasoning). 2. Organizing information (summarizing the data in a table, diagram, or any other representation for comparisons and identifying similarities/differences between the representations- Conceptual Understanding). 3. Processing information (figuring the calculations correctly, writing the solution processes, and provide an appropriate solution to the required task- Procedural Fluency). 4. Making justifications for the suggested solution (giving reasons, providing evidence, and justifying the suggestion- Mathematical Reasoning). Example 1. If a student’s final response is “I suggest buying the mixed fruit juice from Ali’s shop because both volumes are same and Ali’s price is cheaper than Ahmad’s price”. The student has given a very brief answer and the scoring will be as follows: Reference to all data: The student refers to the prices and volumes but he does not refer to the quality-Score 2. Organizing information: The student does not use any representation to present his calculation or conclusion-Score 0. Processing information: The calculations are correct but the student does not write explicitly the solution process-Score 2.

112

Making justifications: The student explains his suggestion, but he does not justify his reasoning (how Ali’s price is cheaper than Ahmad’s price)-Score 2. The total score: 6 Example 2: If a students’ final response is “I suggest buying the mixed fruit juice from Ahmad’s shop. Although the two volumes are same (each fraction in the first offer is equivalent to the each fraction in the second offer), the components of Ahmad’s juice

are 100% fruit juice, but Ali’s juice contains

6 litters of water. So although Ali’s price 8

is cheaper by

1 1 1 1 ( = ) dinar than Ahmad’s price; the better buy is Ahmad’s 4 2 4 4

juice because its quality is better than Ali’s juice. The student has summarized all relevant and irrelevant data in table and given a thorough explanation. The scoring will be as follows: Reference to all data: The student refers to the prices, volumes, and each juice components (quality)-Score 3. Organizing information: The student summarizes all data in a table and provides written explanations.-Score 3. Processing information: The calculations are correct and the student writes explicitly the solution process-Score 3. Making justifications: The student explains his suggestion and justifies his reasoning (how Ali’s price is cheaper than Ahmad’s price)-Score 3.

113

The total score: 12 3.5.2.3 The Metacognitive Knowledge Questionnaire The metacognitive knowledge questionnaire (see appendix E) was adapted from the study of Montague and Bos (1990), assessed students’ metacognitive knowledge regarding their problem-solving strategies, and from Xun (2001) self-report questionnaire. Cronbach’s alpha reliability coefficient of those questionnaires were .83, .86 respectively. The adapted metacognitive knowledge questionnaire consists of 15 items grouped into three categories. The first category (5 items) was focused on strategies used before the solution process (planning) (e.g., “I tried to understand the problem before I attempted to solve it”); the second (5 items) category was focused on strategies used during the solution process (monitoring) (e.g., “I summarized what were given and what were wanted in a table”); and the third (5 items) was focused on strategies used at the end of the solution process (evaluation) (e.g., I tried to find evidence to justify and support my solutions”). Metacognitive questionnaire scoring: Each item was constructed on a 3-point, Likerttype scale ranging from 1 (never) to 3 (always) and a total mean score ranging from 1 to 3. 3.5.3 Materials and Instruments Validity Although the materials and instruments used in this study were derived from theories principles and standards, after the translation to Arabic language, two experienced mathematics teachers, two education mathematics supervisors, and two mathematics education university lecturers in Jordan reviewed the lesson plans, the metacognitive questions card, the scoring procedure of assessing mathematical reasoning items, and

114

the scoring rubrics of assessing the real-life problem. The researcher met the evaluators and discussed the questions regarding these materials and instruments during the evaluation process. The evaluators’ suggestions, feedback, and comments were taken into account until there were no discrepancies among the evaluators. Then the evaluators reviewed the mathematics achievement test questions and the metacognitive knowledge questionnaire items. Each looked at each question in the test and at each item in the questionnaire and assessed which of the mathematical proficiency strand (CU, PF, or MR) the question represented and which of the metacognitive knowledge component (planning, monitoring, or evaluation) the item represented, and rated their confidence in their response, using scale from 1 (weak) to 10 (strong). Only questions and items, which had received 7 or more scores from all evaluators, were selected as test questions and questionnaire items following Chung (2002). Evaluators agreed that the questions 1, 2.1 ,2.2, 3.1, 3.2, 6, 7, 8, 9.1, 9.2, 10.1, and 10.2, were represented to CU, CU, MR, CU, MR, PF, PF, PF, CU, MR, CU, and MR strands respectively, with all reporting confidence scores 10, questions 4.1(a, b, c, d), 4.2 (a, b, c, d), and 5 (a, b, c, d) were represented to CU, MR, and PF strands respectively, with all 9 scales, and question 11 (real-life problem) (criterion 1 to 4) were represented to MR, CU, PF, and MR strands respectively, with all reporting confidence scores 8. Since evaluators were in disagreement about question 4.1 (e), 4.2 (e), 5 (e), and 11 (criterion 5), the questions were removed from the test. For the questionnaire items validity, evaluators agreed that the first five items of each scale (planning, monitoring, and evaluation) were represented, with all reporting confidence scores 10. However, there were disagreement about the last three items of each scale (9 items), therefore they were removed from the questionnaire. After an overall

115

agreement was reached on the validity of the materials and instruments, they were considered valid materials and instruments for conducting this study. 3.5.4 Instruments Reliability Two major instruments were used in this study i.e., the mathematics achievement test and the metacognitive knowledge questionnaire. Although the instruments used in this study were adapted from reliable instruments, with the additional items and translation to Arabic, a pilot test was carried out and the scores from the pilot study test and the metacognitive questionnaire were collected and a set of reliability tests were conducted to determine the Cronbach’s Alpha reliability coefficients. Cronbach’s alpha reliability coefficient of the mathematics achievement test was .88, and it was .84 for the metacognitive knowledge questionnaire. Cronbach’s Alpha reliability coefficients for the metacognitive questionnaire categories were .64, .66, .60 for planning, monitoring, and evaluation respectively. The Cronbach’s Alpha reliability coefficients showed that the study instruments were satisfactory reliable.

3.6 Procedures Prior to the implementation of the study, the researcher obtained permissions from a number of different parties for conducting the pre-experimental study and the experimental study. Permissions were sought from the educational development and research department of the Jordan Ministry of Education (see appendix F), the First Public Educational Directorate in Irbid Governorate (see appendix G) where the participating schools are located, and from the participating schools’ principals.

116

3.6.1 The Pilot Study Prior to the formal study sessions, a pilot study was conducted to validate research procedures. The researcher selected randomly 80 participants from a randomly selected primary school, who were not going to participate in the formal study. There were two purposes to the pilot study: first, to test the materials and instruments, in terms of using the metacognitive questions cards, sessions duration, training of teachers, and the test and the questionnaire durations; and secondly, to test the instruments reliability i.e., the mathematics achievement test and the metacognitive knowledge questionnaire. Two male teachers who had a similar level of education (B.Ed. major in mathematics) and had more than 7 years of experience in teaching mathematics were selected to teach the participants in the pilot study. The teachers were exposed to one week training about teaching adding and subtracting fractions with MSCL and CL methods. The participants were randomly assigned to the two experimental conditions i.e., MSCL and CL groups. Within each condition, teachers conducted classes according to their assigned teaching methods for 14 sessions. At session 15, all participants were administered the mathematics achievement test and immediately responded to the metacognitive questionnaire items. 3.6.2 The Formal Study For the experiment, the researcher randomly selected three schools from the 44 male primary schools in Irbid Governorate. Permission was sought from each school’s principal. Three mathematics teachers with a similar level of education (B.Ed. major in mathematics), had more than 7 years of experience of teaching mathematics in heterogeneous classrooms were selected (one from each school). Each teacher taught two classes in each school. Each teacher’s classes were randomly assigned into the

117

three instructional methods described earlier. The researcher then discussed with each teacher about his assigned instructional method and appointed one day with each one to discuss about training. 3.6.3 Groups’ Equivalence To test the assumption that the participants across the three groups were equivalent, the pre-test was conducted two months before the beginning of the study. The pre-test was focused on students’ conceptual understanding, procedural fluency, and mathematical reasoning. The pre-test papers were scored by the researcher. To determine if there were statistically significant differences between the groups’ mean scores i.e., the high-ability students and the low-ability students, the scores by the three groups were entered into the Statistical Package for Social Science (SPSS) for Windows computer software (version 11.5). 3.6.4 Teachers’ Training Prior to the beginning of this study, the teachers assigned to the experimental groups participated in one week training sessions that focused on pedagogical issues regarding teaching mathematics. The teachers were informed that they would be part of an experiment in which new instructional methods were being tested. They worked with the new methods and materials and learned how to use them with their students. The materials included the mathematics textbooks, explicit lesson plans, and examples of metacognitive questions. Within each school, the teachers continued conducting classes according to their assigned teaching methods until the end of the first semester. In the present study, the focus was on the “Adding and Subtracting Fractions” unit that was taught in all classrooms for 14 sessions at the end of the first semester.

118

The CLMS teacher was trained explicitly about using cooperative learning with metacognitive scaffolding in the teaching of Adding and Subtracting Fractions. He was exposed to some examples about the nature of the metacognitive questions and how to use and train students to use them in a cooperative learning setting. He was informed to use metacognitive questions in his explanations and coach his students to use metacognitive questions when they solve the mathematical problems. The procedures of selecting groups and assigning group members were explained to the teacher. The researcher met the teacher for feedback and assessment regarding the application of the teaching method. The CL teacher was trained about teaching mathematics within cooperative learning setting, and about selecting groups and assigning groups’ members. He was not exposed to any training about metacognitive scaffolding method. The researcher met the teacher for feedback and assessment regarding the application of the teaching method. Finally, the T teacher was not exposed to the metacognitive scaffolding or to the cooperative learning training, he was asked to teach as he used to teach in a usual manner. The researcher checked his lesson plans and his methods of teaching to ensure that he followed the traditional method.

119

3.6.5 Implementation of the Study 3.6.5.1 The Cooperative Learning with Metacognitive Scaffolding (CLMS) Method In this treatment, the pre-test was conducted first and then students were informed that in the following weeks they would be exposed to an instructional method that would help them become more effective managers of their own learning activities. The teacher introduced the processes of cooperative learning with metacognitive scaffolding method to the students. He discussed with them about the importance and the role of this method in developing their mathematics performance, reasoning, and metacognitive knowledge. The teacher spent some time on explicitly introducing the concepts of how students can become metacognitive students within this learning environment, why they would learn metacognitive strategies, and how they could apply these strategies in solving real-life problems. After the discussion on cooperative learning with metacognitive scaffolding method, students were assigned into groups based on their ability. They were divided into high and low-abilities based on their pre-test scores in mathematics performance and mathematical reasoning. The median of the scores was the criterion of assigning students to the group. Because students’ scores were interval variables, they were converted to nominal variables. The scores were placed in numerical order and then the median score was located. Scores above the median (16) were labeled as high-ability and below the median were labeled as low-ability following Tuckman (1999). Each group was formed by randomly choosing two high-ability students and two low-ability students. The remaining groups were selected by repeating the same procedure with the reduced list.

120

When the grouping was completed, students’ roles in their group were assigned. Within each group, the teacher randomly assigned a metacognitive question asker, summarizer, recorder, and presenter and then he described each role. The metacognitive questions asker read the questions from the metacognitive questions card and asked his group members. The summarizer summarized orally the main ideas and the key points to solve the problem, and the recorder wrote down the solution steps, the explanations, and the justifications of that solution. Finally, the presenter presented, explained, and justified the solution to the whole class. These roles were rotated among students after each session so that each group member played each one several times. The teacher applied the CLMS instructional method two months before the formal experiment with practice units. For the formal experiment, just before the “Adding and Subtracting Fractions” unit was taught, students were informed that at the end of this unit, they would be asked to complete a mathematics achievement test and a questionnaire. The formal experiment lasted 15 sessions (14 sessions for implementing the method and 1 session for administrating the test and the questionnaire). In the first session, the teacher introduced and explained the new topic for about 30 minutes to the whole class by asking him-self metacognitive questions regarding planning, monitoring, and evaluation. For example, before solving the problem, instead of saying, First we..., next we..., then we..., the teacher said, “I need to know what the whole task is about, is it about the whole numbers, fractions, additions, or subtraction, etc? What is given and what is not given? What in my prior knowledge will help me with this particular task? Last time I have learned about adding fractions with the same denominators, but this task includes fractions with different denominators, so what should I do? Do I know where I can go to get some

121

information on this task? What are some strategies that I can use to learn this? I should find a way to transform one of the denominators to be same as the other then I can add, so how should I do this?” The teacher then coached and encouraged students to ask these metacognitive questions within the cooperative setting. Students were encouraged to talk about the task, explain to each other, and represent it from different perspectives. During his explanation process, the teacher also asked metacognitive questions regarding monitoring. For example, did I understand what I have just decided to do? Am I on the right track? How can I spot an error if I make one? How should I revise my plan if it is not working? Am I keeping good notes or records? Again, students were encouraged and trained to ask these questions. At the end of his explanation, the teacher asked and trained students to ask metacognitive questions regarding evaluation such as: Did the solution make a sense, and how can I decide that? Did my particular strategy produce what I had expected? What could I have done differently? How might I apply this line of thinking to other problems? Finally, the teacher summarized the learning processes through metacognitive questions before (planning), during (monitoring), and after (evaluation) of the learning task and encouraged students to apply them in learning adding and subtracting fractions. After the teacher’s explanation, the metacognitive questions cards were distributed to the groups. The students were asked to do their exercises and solve the assigned mathematical problems in groups for about 15 minutes. The teacher had explained to the students about the reasons for doing each of the steps on the matacognitive questions cards. This is important because according to Palincsar and Brown (1984), providing reasons for doing a particular action (i.e., responding to the “why do we do

122

this” question) during a learning strategy usage training will increase the likelihood that the strategy will continue to be used by the participating students after the training. In this way, the metacognitive questions asker had read the problem and asked aloud, the colleagues listened to the mathematics metacognitive question and tried to answer. Whenever there was no consensus, the group members discussed the issue until the disagreement was resolved. When the disagreement was resolved, the summarizer orally summarized the solution, the explanation, and the justification and discussed with his colleagues. With the solution, explanation, and justification were in hand, the recorder has written them down and the presenter has presented to the whole class. During these processes, the teacher monitored each learning group and intervened by asking more metacognitive questions if necessary. At the end of the session, the teacher collected the metacognitive questions cards and assessed and evaluated students’ performance, discussed with the whole class to ensure that students carefully process the effectiveness of their learning group, and had students celebrate the work of group members. For the next sessions, the teacher and students followed the same method and procedures and the group members’ roles were rotated after each session. However, the metacognitive scaffolding input by the teacher was gradually reduced, for example, the teacher’s time in the first session was 30 minutes, in the second session it was about 25 minutes, in the third session it was about 20 minute and so on until the time became when the teacher taught for about 10 minutes regarding the new topic and the students continued learning by their own using the metacognitive questions cards. After one month of implementing the CLMS instructional method, namely in

123

the last mathematics session of this experiment (session 15), the students were asked to complete the mathematics achievement test. After completing the test, they were immediately asked to complete the metacognitive questionnaire. 3.6.5.2 The Cooperative Learning (CL) Method In this treatment, students taught via cooperative learning instructional method with no metacognitive scaffolding. The pre-test was conducted first and then students were informed that in the following weeks they would be exposed to an instructional method that would help them to improve their learning activities. The teacher introduced the cooperative learning method stages and discussed with the students about the importance of using this method in mathematics classroom. Students were assigned into heterogeneous small groups following the same procedures of assigning students to the groups in the CLMS condition. Because there were students left over, one group of three members was formed (one high-ability student and two low-ability students). Within each group, the teacher randomly assigned reader, summarizer, recorder, and presenter and then he has described each role. The teacher applied the CL instructional method two months before the formal experiment with practice units. For the formal experiment, just before the “Adding and Subtracting Fractions” unit was taught, students were informed that at the end of this unit, they would be asked to complete a mathematics achievement test and a questionnaire. The formal experiment lasted 15 sessions (14 sessions for implementing the method and 1 session for administrating the test and the questionnaire). In the first session, the teacher introduced and explained the new topic for 25 minutes to the whole class and then proceeded to teach in a usual manner. For example, he used the board and explained the main ideas of today’s lesson. After the

124

teacher’s explanation of the new topic to the whole class, students were asked to do their exercises and solve the assigned mathematical problems in groups for 20 minutes. The reader read the problem aloud; the colleagues discussed about the learning task and asked themselves different questions (but they were not trained to ask metacognitive questions). The summarizer, the recorder, and the presenter played the same roles of their counterparts in the CLMS group. At the end of the session, the students ensured that all of them mastered the task. During the session, the teacher intervened when needed to improve task work and teamwork, but he did not use metacognitive scaffolding, namely, he asked questions regarding the task such as: what are the procedures of adding two fractions with different denominators, and he responded to students’ questions. Finally, the teacher assessed and evaluated students’ performance, ensured that students carefully process the effectiveness of their learning group, and had students celebrate the work of group members. For the next sessions, the teacher and students followed the same method and procedures and the group members’ roles were rotated each session. After one month, namely in the last mathematics session of this experiment (session 15), the students were asked to complete the mathematics achievement test. After completing the test, they were immediately asked to complete the metacognitive questionnaire. 3.6.5.3 The Traditional (T) Method The control group served as a comparison group with no intervention. Therefore, the teacher of this group continued teaching as he usually did, and the students were not exposed to cooperative learning or metacognitive scaffolding. The pre-test was conducted two months before teaching the “Adding and Subtracting Fractions” unit. Just before the “Adding and Subtracting Fractions” unit was taught, students were

125

informed that at the end of this unit, they would be asked to complete a mathematics achievement test and a questionnaire. In this condition, the “Adding and Subtracting Fractions” teaching lasted 15 sessions (14 sessions for teaching and 1 session for administrating the test and the questionnaire). In the whole 14 sessions of implementing T method, the teacher introduced, explained, and manipulated the new concepts and procedures of today’s lesson using the board and the textbook for 35 minutes to the whole class. After the teacher’s explanation, the students practiced the mathematical items individually using their textbooks and teacher’s notes and sometimes employed any method the teacher saw fit for 10 minutes. When the students faced difficulties during solving the mathematical problems, and finally could not find the solution, they asked for the teacher’s help. So the teacher intervened when needed to help some students to solve their mathematical problems. Sometimes the teacher explained and informed the students about the procedures of solving the problem. At the end of each session, the teacher reviewed the day’s lesson with the whole class. In session 15, the students were asked to complete the mathematics achievement test. The metacognitive questionnaire was passed out to the students immediately after they completed the mathematics achievement test. 3.6.5.4 Monitoring the Implementation of the Study During the first two months of implementing this study, three mathematics education supervisors, whose job was to regularly visit the three teachers in their classes, visited the three teachers twice a month. Each mathematics education supervisor was informed to observe his assigned teacher following the checklists prepared by the researcher to ensure the fidelity to the implementation. The checklist of the CLMS group contained questions such as: Did the teacher follow his lesson plans correctly? Did the teacher ask metacognitive questions during his explanations? Did the teacher

126

assign the groups correctly? Did the teacher gradually reduce his metacognitive scaffolding input? Did the teacher distribute the metacognitive questions cards to the all groups? Did each group member play different roles? The checklist of the CL group contained questions such as: Did the teacher follow his lesson plans correctly? Did the teacher assign the groups correctly? How long did the teacher's explanation last? How long did the students work cooperatively? Did each group member play different roles? The checklist of the T group contained questions such as: Did the teacher follow his lesson plans correctly? How long did the teacher's explanations last? How long did student spend to solve the mathematics problems individually? During the last month of implementing this study, namely, during the teaching of "Adding and Subtracting Fractions Unit", the three mathematics education supervisors visited the three teachers twice a week and followed the same checklists to ensure the implementation fidelity. Also the researcher met each teacher twice a week to ensure fidelity to the treatment following the checklists used by the three mathematics education supervisors. At the end of session 15, the researcher collected the mathematics test and the metacognitive questionnaire papers from the three participating groups. The mathematics test items and the metacognitive questionnaire items were scored by the researcher using the scoring rubrics.

127

3.7 Data Analysis Procedure and Method 3.7.1 The pre-Experimental Study Findings Analysis Groups’ Equivalence To test the assumption that the participants (high and low ability students) across the three groups are equivalent, the participants’ average scores on the pre-test in mathematics performance and mathematical reasoning(mathematics and Arabic) will be were analyzed to determine if there are were statistically significant differences between the groups’ mean scores. Since there are were two dependent variables i.e., mathematics performance (MP) and mathematical reasoning (MR), and a three groups and one moderator variable with two levels (high-ability and low-ability), two-way multivariate analysis of variance (two-way MANOVA) test willstatistical technique was conducted. In addition, a reliability test was conducted for the mathematics performance and mathematical reasoning test and for the metacognitive questionnaire items to determine the Chronbach Alpha reliability values.

be used to compare the three mean scores, namely, two-way ANOVA will be used to compare the high-achievers across the three group and to compare the low-achievers across the groups.

3.7.2 Instruments Reliability

The test-retest reliability coefficient for the mathematical achievement test and the metacognitive questionnaire will be measured through entering the findings into the SPSS computer program. The correlation coefficient will be measured to ascertain the instruments reliability.

3.7.2 The Experimental Study Findings Analysis

128

At the end of this study, the two experimental groups and the control group participants will completed the mathematics achievement performance and mathematical reasoning test and filled out the metacognitive questionnaire. The test and the questionnaire items were scored by the researcher and The results will be analyzed to determine if there are were any statistically significantce differences between the three groups on the dependent variables. While there arewas an

independent variable with (three levels, a moderator variable with two levels,) and three dependent variables, and the pre-test as a covariate, two-way multivariate analysis of variancecovariance (two-way MANCOVA) test will bewas conducted to compare the three adjusted mean scores on mathematics achievementperformance (MP), mathematical reasoning (MR), and metacognitive knowledge (MK).

MANCOVA will be was conducted first to compare MP, MR, and MK of the three groups. Then MANOVA was conducted with splitting file technique to compare highability students against high-ability students’ MP, MR, and MK across the three groups. The same technique was used to compare low-ability low-achievers’ students against low-ability students’ MP, MR, and MK across the three groups. Because the overall two-way MANCOVA results were statistically significant, a series follow up two-way analysis of covariance (two-way ANCOVAs) were used to identify where the differences resided. Since the follow up ANCOVAs results were statistically significant, the post hoc pair wise comparison technique using the /lmatrix command was used to identify where the differences in adjusted means resided. Finally, by conducting two-way MANCOVA without splitting files, the interaction effects between the instructional method and the ability level (high-ability and low-ability) was measured. All of the statistical analysis tests will bewere computed at 0.05 level of significance.

129

3.7.3 Justifications for using two-way MANCOVA / MANOVA Two-way multivariate analysis of covariance is used to determine how each dependent variable is influenced by two independent variables while controlling for a covariate (Hair et al., 1998). MANCOVA is to reduce the size of the error term in the analyses thereby increasing power (Stevens, 1986). Analysis of covariance adjusts the mean of each dependent variable to what they would be if all groups started out equally on the covariate. Analysis of covariance gives results preferable to those of a direct comparison of gain scores i.e., post-test minus pre-test for the two groups, because gains are limited in size by the difference between the test’s ceiling and the magnitude of the pre-test score (Tuckman, 1999). In this study, pre-MP and pre-MR have been shown to correlate with the dependent variables, thus they were considered as appropriate covariates. Two-way MANOVA is used to examine the effects of two or more independent variables on a set of dependent variables (Stevens, 1986). A two-way MANOVA enables us to (1) examine the joint effect of the independent variables on the dependent variables, and (2) get more powerful tests by reducing error (within-cell) variance (Stevens, 1986). A moderate to strong correlation among the dependent variables is an additional justification for using two-way MANOVA. If subsequent overall MANOVA results are statistically significant, a one-way analysis (ANOVA) is conducted to further examine or identify where the differences reside. If there is no correlation, or if the correlation is weak among the dependent variables, MANOVA is not considered since

130

a single outcome measure may be diluted in a joint test involving many variables that display no effect. In such a situation, individual univariate tests are directly conducted. 3.7.4 Pearson’s Correlation The scores of the mathematics performance and mathematical reasoning test and the scores of the metacognitive questionnaire were analyzed by examining the relationships among the multiple dependent variables by using Pearson's correlation technique. The purpose was to determine if there were statistical justifications to use multivariate analysis of variance (MANOVA) (Stevens, 1986). The results of the Pearson's correlation (see Table 3.3) indicated an overall correlation among the three dependent variables (mathematics performance MP, mathematical reasoning MR, and metacognitive knowledge MK), significant at the .01 level. Table 3.3 Pearson’s correlation among the three dependent variables (MP, MR, and MK)

Variable MP MR MK MP

Participants (n = 240)

MR

MK

_ .746** .652** _ .757** _

Note. ** suggests that correlation is significant at the 0.01 level (2-tailed). 3.7.5 Assumptions for MANOVA / MANCOVA A preliminary analysis was conducted to determine whether the prerequisite assumptions of MANOVA / MANCOVA were met before proceeding the multivariate analysis. Thus, the assumption of normality, equality of variance-covariance matrices,

131

and the linear relationship between the covariates and the dependent variables were examined. The assumption of normality was supported by the data. The M-estimators had strong agreement among 4 estimators. All Q-Q plots fall along the straight line showing that the normality in all variables was reasonable. Box’s M Test of Equality of Covariance Matrices tests the null hypothesis that the observed covariance matrices of the dependent variables are equal across groups. The Levene’s Test tests the null hypothesis that the error variance of the dependent variable is statistically similar across groups. The value for Box’s M of comparing the three groups regardless of the ability level = 68.217, F (30, 122811) = 2.200, p < .001 was significant, thus rejecting the assumption of homogeneity of the variances. However, rejecting this assumption has minimal impact if the groups are of approximately equal size i.e., if the largest group size divided by the smallest group size is less than 1.5 (Hair et al., 1998). Therefore, for this particular study, rejecting of this assumption has minimal impact since the groups were of approximately equal sizes. The value for Box’s M of comparing high-ability students across the three groups = 18.185, F (12, 65332) = 1.459, p > .001 was not significant. In addition, the value for Box’s M of comparing low-ability students across the three groups = 29.420, F (12, 66085) = 2.360, p > .001 was not significant, thus accepting the assumption of homogeneity of the variance. The results from the Levene’s Test for homogeneity of variance of comparing the three groups regardless of the ability level for each of the dependent variables

132

indicated that homogeneity of variance has been met for all the three dependent variables. For MP, F (5, 234) = 3.210, p > .001, for MR, F (5, 234) = 1.681, p > .001, and for MK, F (5, 234) = 4.224, p >.001. The results from the Levene’s Test for homogeneity of variance of comparing the high-ability students across the three groups for each of the dependent variables indicated that homogeneity of variance has been met for all the three dependent variables. For MP, F (2, 117) = 2.801, p > .001, for MR, F (2, 117) = .427, p > .001, and for MK, F (2, 117) = 1.955, p > .001. In addition, the results from the Levene’s Test for homogeneity of variance of comparing the low-ability students across the three groups for each of the dependent variables indicated that homogeneity of variance has been met for all the three dependent variables. For MP, F (2, 117) = .339, p > .001, for MR, F (2, 217) = .042, p > .001, and for MK, F (2, 117) = 4.378, p > .001. To examine the assumption that the covariates must have some relationship with the dependent variables (Hair et al., 1998), the scores of the mathematics performance and mathematical reasoning test and the scores of the metacognitive questionnaire were analyzed by examining the relationships among the covariates and the dependent variables by using Pearson’s correlation technique. The results of the Pearson’s

correlation (see Table 3.4) indicated an overall correlation among the two covariates (pre-MP and pre-MR) and the three dependent variables (mathematics performance MP, mathematical reasoning MR, and metacognitive knowledge MK), significant at the .01 level. Table 3.4 Pearson’s correlation among the covariates (pre-MP and pre-MR) and the dependent variables (MP, MR, and MK)

133

Covariate / Variable Pre-MP Pre-MR

MP

Participants (n = 240)

MR .513** .658**

MK .316** .442**

.522** .719**

Note. ** suggests that correlation is significant at the 0.01 level (2-tailed). After determining that the assumptions were met, the multivariate statistical output was examined. Then, providing the MANCOVA result was statistically significant, the univariate results were examined for each dependent variable. For the significant univariate results, the post hoc comparisons were performed to identify where the differences resided. The pairwise comparisons statistic was used for the post hoc results. The results of the multivariate tests, the univariate tests, the pairwise comparisons among the three dependent variables, the interaction effect, as well as the descriptive statistics for the dependent variables are reported in Chapter Four. CHAPTER FOUR RESULTS 4.1 Introduction This chapter presents the results of the study from the data analyses of the preexperimental study as well as the experimental study. The analyses were carried out through various statistical techniques such as the two-way multivariate analysis of variance (MANOVA), the univariate analysis of variance (ANOVA), the two-way multivariate analysis of covariance (two-way MANCOVA), the one-way multivariate analysis of covariance (one-way MANCOVA), the two-way analysis of covariance (two-way ANCOVA), and the post hoc pair wise comparison using the /lmatrix

134

command analysis. The data were compiled and analyzed using the Statistical Package for the Social Science (SPSS) for Windows computer software (version 11.5). The results of the pre-experimental study, in response to the groups’ equivalence are reported first. Hypotheses regarding the effects of the instructional methods on students’ mathematics performance (MP), mathematical reasoning (MR), and metacognitive knowledge (MK) are tested, and the findings of testing these hypotheses are presented. Next the hypotheses regarding the effects of the instructional methods on high-ability and low-ability students’ MP, MR, and MK are tested, and the findings of testing these hypotheses are presented. Each hypotheses tested is followed by a summary of testing that hypotheses. Finally, the summary of findings to research questions 1 - 4 is presented.

4.2 The pre-Experimental Study Results The purpose of the pre-experimental study was to test the assumption that the participants across the three groups were equivalent in mathematics performance and mathematical reasoning. To achieve this purpose, a pre-test that measures premathematics performance and pre-mathematical reasoning was conducted before the beginning of the study. While there were three groups with moderator variable with two levels i.e., high-ability and low-ability, and two dependent variables i.e., premathematics performance (pre-MP) and pre-mathematical reasoning (pre-MR), twoway multivariate analysis of variance (MANOVA) with splitting file technique was conducted to determine if there were statistically significant differences between the groups’ mean scores i.e., high-ability against high-ability students and low-ability against low-ability students across the three groups.

135

4.2.1 Statistical Data Analysis Table 4.1 summarizes the descriptive statistics for the dependent variables (pre-MP and pre-MR) by the groups. Both dependent variables had the same points (22 points for each). The scores of high-ability student on pre-MP across the three groups had relatively similar means, 11.1750, 11.7368, and 11.0476 for CLMS, CL, and T respectively. The scores of high-ability student on pre-MR had also relatively similar means, 7.5000, 7.9474, and 7.9762 for CLMS, CL, and T respectively. For low-ability students, the scores of the three groups on pre-MP were very close, (8.5500, 9.1220, and 9.2564 for CLMS, CL, and T respectively). The scores of the three groups on preMR were very close, (3.2750, 2.9756, and 3.4872 for CLMS, CL, and T respectively).

**Table 4.1 Means and standard deviations on each dependent variable (pre-MP and preMR), by the groups
**

Dependent Variables Ability High (H) Low (L) H (n = 40) L (n = 40) CL H (n = 38) L (n = 41) T H (n = 42) L (n = 39) Mean SD Mean SD Mean SD Mean SD Mean SD Mean Pre-MP Pre- MR

Group CLMS

11.1750 1.7525 8.5500 1.0857 11.7368 2.0754 9.1220 1.1289 11.0476 1.5134 9.2564

7.5000 .5991 3.2750 1.5684 7.9474 1.1377 2.9756 1.4639 7.9762 1.8144 3.4872

136

SD

1.5706

1.4346

Note.

Total score on pre-MP = 22, and total score on pre-MR = 22

To examine if there were significant statistical differences between the high-ability students on pre-MP and pre-MR across the three groups, and if there were significant statistical differences between the low-ability students on pre-MP and pre-MR across the three groups, two-way multivariate analysis of variance (MANOVA) was conducted. Table 4.2 presents the results of two-way multivariate analysis of variance, showing overall differences between high-ability students and low-ability students across the three groups on pre-MP and pre-MR. To evaluate the multivariate (MANOVA) differences, Pillai’s Trace criterion was considered to have acceptable power and to be the most robust statistic against violations of assumptions (Coakes and Steed, 2001). The MANOVA results of comparing high-ability students against high-ability students and low-ability students against low-ability students across the three groups were statistically not significant (F = 1.773, p = .135), (F = 2.255, p = .064) respectively. Further, the results of the univariate ANOVA tests, which are represented in Table 4.2, indicated that there were no significant statistical differences between the high-ability students in pre-MP and pre-MR, with an F ratio (2, 117) of 1.653 ( p = .196) and 1.700 ( p =.187) respectively. Also the results indicated that there were no significant statistical differences between the low-ability students in pre-MP and pre-MR, with an F ratio (2, 117) of 2.803 ( p = .65) and 1.625 ( p = .201) respectively. This means that there were no statistically significant differences between high-ability students and no statistically significant differences between low-ability students across the three groups in pre-MP and pre-MR. Therefore, the assumption that the high-ability

137

participants across the three groups and the low-ability participants across the three groups are equivalent in mathematics performance and mathematical reasoning was met.

Table 4.2 Summary of multivariate analysis of variance (MANOVA) pre-MP and pre-MR results and follow-up analysis of variance (ANOVA) results.

MANOVA Effect and Dependent Variables Group Effect Highability Pre-Mathematics Performance (pre-MP) Pre-Mathematical Reasoning (pre-MR) Group Effect Lowability Pre-Mathematics Performance (pre-MP) Pre-Mathematical Reasoning (pre-MR) Pillai's Trace 2.255 ( p =.064) 2.803 ( p = .065) 1.625 ( p = .201) Multivariate F Pillai's Trace 1.773 ( p =.135) 1.653 ( p =.196) 1.700 ( p =.187) Univariate F df = 2, 117

138

4.3 The Experimental Study Results The purpose of the experimental study was to examine the effects of the instructional methods on mathematics performance (MP), mathematical reasoning (MR), and metacognitive knowledge (MK), specifically on high-ability and low-ability students’ mathematics performance, mathematical reasoning, and metacognitive knowledge while controlling students’ pre-MP and pre-MR on the pre-test. A two-way multivariate analysis of covariance (MANCOVA) was conducted to analyze the effects of the instructional method on the three dependent variables, as well as the interaction between the instructional method and the ability levels effects on the three dependent variables. The statistical differences of the three groups were compared and analyzed according to each of the three dependent variables. The research hypotheses were tested using

139

the results from the two-way multivariate analysis of covariance (MANCOVA) and univariate analysis of covariance (ANCOVA). The results of the analysis were used to answer Research Questions 1-4. 4.3.1 Testing of Hypothesis 1 Students taught via cooperative learning with metacognitive scaffolding (CLMS) instructional method will perform higher than students taught via cooperative learning (CL) instructional method who, in turn, will perform higher than students taught via traditional (T) instructional method in (a) mathematics performance (MP), (b) mathematical reasoning (MR) and (c) metacognitive knowledge. Table 4.3 presents overall means, standard deviations, adjusted means, and standard errors of each dependent variable by the instructional method, CLMS, CL, and T. Table 4.3 Means, standard deviations, adjusted means and standard errors for each dependent variable by the instructional method

Dependent Variables Mathematics Performance (MP) Mean SD Adj. mean Std. Error Mathematical reasoning (MR) Mean SD Adj. mean Std. Error

The Instructional Method CLMS CL T N= 80 N= 79 N= 81 18.6500 17.5570 16.7654 2.3390 18.742 .156 16.1500 2.2842 16.289 .146

a a

**2.7351 17.611 .157 14.1646 2.7336 14.184 .147
**

a a

**2.2928 16.639 .155 12.7284 2.3875 12.576 .145
**

a a

140

Metacognitive Knowledge (MK)

Mean SD Adj. mean Std. Error

**2.2975 .2541 2.299 .021
**

a

**1.9485 .3094 1.954 .021
**

a

**1.7243 .2788 1.718 .020
**

a

**Note. a. Evaluated at covariates appeared in the model: pre-MP = 10.1417, preMR = 5.5250.
**

Total score on MP = 22, total score on MR = 22, and total score on MK = 05

To examine if there were statistically significant differences in mathematics performance, mathematical reasoning, and metacognitive knowledge adjusted mean scores between the CLMS, the CL, and the T groups, while controlling the pre-MP and the pre-MR, multivariate analysis of covariance (MANCOVA) was conducted.

Table 4.4 presents the results of multivariate analysis of covariance (MANCOVA), showing overall differences for the independent variable of instructional method effect and the three dependent variables, while controlling pre-MP and pre-MR. The Pillai’s Trace was used to evaluate the multivariate (MANCOVA) differences. The MANCOVA results of comparing the three groups were statistically significant (F = 46.575, p = .000). The covariates pre-MP (F = 15.020, p = .000) and pre-MR (F = 16.553, p = .000) had significant effects. This means that there were some statistical differences on at least one dependent variable. Further, the results of the univariate ANCOVA tests, which are represented in table 4.4, indicated that there were statistically significant differences in the three dependent variables (MP, MR, and MK). The F ratio of MP (2, 237) was 45.600 ( p = .000). This means that the instructional method had a main effect on MP. This effect accounted

141

for 28% of the variance of MP (Eta2 = .282). The F ratio of MR (2, 237) was 162.490 ( p = .000). This means that the instructional method had a main effect on MR. This effect accounted for 58% of the variance of MR (Eta 2 = .583). The F ratio of MK (2, 237) was 202.729 ( p = .000). This means that the instructional method had a main effect on MK. This effect accounted for 64% of the variance of MK (Eta2 = .636).

Table 4.4 Summary of multivariate analysis of covariance (MANCOVA) results by the instructional method and follow-up analysis of variance (ANOVA) results.

MANCOVA Effect, Dependent Variables, and Covariate Group Effect Mathematics Performance(MP) Mathematical Reasoning(MR) Metacognitive knowledge (MK) Pre-MP Pre-MR 15.020 ( p = .000) 16.553 ( p = .000) Multivariate F Pillai's Trace 46.575 ( p = .000) 45.600 ( p = .000) 162.490 ( p = .000) 202.729 ( p = .000) Univariate F df = 2, 237

142

The MANCOVA results of comparing the three groups on the three dependent variables indicated that there were statistically significant differences between at least two groups in the three dependent variables. Therefore, the researcher further investigated the univariate statistics results (analysis of covariance ANCOVA) by performing a post hoc pairwise comparison using the /lmatrix command for each dependent variable in order to identify significantly where the differences in the adjusted means resided. Table 4.5 is a summary of post hoc pairwise comparisons. .

**Table 4.5 Summary of post hoc pairwise comparisons
**

Dependent Variable

Mathematics Performance (MP) Comparison Group CLMS vs. CL CLMS vs. T CL vs. T Adj.Mean Difference 1.131 Sig .000 Mathematical Reasoning (MR) Adj.Mean Difference 2.105 Sig .000 Metacognitive Knowledge (MK) Adj.Mean Difference .345 Sig .000

2.103

.000

3.713

.000

.581

.000

.972

.000

1.608

.000

.236

.000

Note.

143

The adjusted mean differences shown in this table are the subtraction of the second condition (on the lower line) from the first condition (on the upper line); for example, 1.131 (Adjusted Mean Difference for Mathematics Performance) = CLMS – CL.

Table 4.3 displays the means, standard deviations, adjusted means and standard errors of different conditions by the dependent variables. Table 4.4 and table 4.5 show that there are statistical adjusted mean differences among the three conditions in the three dependent variables. The adjusted mean differences are presented below. Mathematics performance. The cooperative learning with metacognitive scaffolding (CLMS) group (Mean = 18.7, SD = 2.3, Adj.mean = 18.7, p = .000) significantly outperformed the other two groups (CL and T), with an adjusted mean difference of 1.131 and 2.103 respectively. On other hand, the cooperative learning (CL) group (Mean = 17.6, SD = 2.7, Adj.mean = 17.6, p = .000) significantly outperformed the control group (T) (Mean = 16.8, SD = 2.3, Adj. mean = 16.6) with an adjusted mean difference of .972. (Effect sizes on MP were .47 and .34 for comparing the CLMS and CL, and CL and the T group, respectively). Mathematical reasoning. The CLMS group (Mean = 16.6, SD = 2.3, Adj.mean = 16.3, p = .000) significantly outperformed the CL and T group, with an adjusted mean difference of 2.105 and 3.713 respectively. The CL group (Mean = 14.7, SD = 2.7, Adj.mean = 14.9, p = .000) significantly outperformed the T group (Mean = 12.7, SD = 2.4, Adj.mean = 12.6) with an adjusted mean difference of 1.608. (Effect sizes on MR were .83 and .60 for comparing the CLMS and CL, and CL and the T group, respectively). Metacognitive knowledge. The CLMS group (Mean = 2.3, SD = .3, Adj.mean = 2.3, p = .000) significantly outperformed the CL and T group, with an adjusted mean

144

difference of .345 and .581 respectively. The CL group (Mean = 1.9, SD = .3, Adj.mean = 2, p = .000) significantly outperformed the T group (Mean = 1.7, SD = .3, Adj.mean = 1.7), with an adjusted mean difference of .236. (Effect sizes on MK were 1.25 and .80 for comparing the CLMS and CL, and CL and the T group, respectively).

4.3.2 Summary of Testing Hypothesis 1 (CLMS > CL > T) The statistical results confirm the hypothesis, showing that students taught via cooperative learning with metacognitive scaffolding instructional method performed significantly higher than the students taught via cooperative learning instructional method who, in turn, performed significantly higher than the students taught via the traditional instructional method in (a) mathematics performance, (b) mathematical reasoning, and (c) metacognitive knowledge. 4.3.3 Testing of Hypotheses 2 High-ability students taught via cooperative learning with metacognitive scaffolding instructional method (CLMSH ) will perform higher than high-ability students taught via cooperative learning instructional method ( CLH ) who, in turn, will perform higher than high-ability students taught via the traditional instructional method (T H ) in (a) mathematics performance (MP), (b) mathematical reasoning (MR) and (c) metacognitive knowledge. Table 4.6 presents overall means, standard deviations, adjusted means, and standard error of each dependent variable for high-ability students by the instructional method, CLMS, CL, and T.

145

Table 4.6 Means, standard deviations, adjusted means and standard errors for each dependent variable for high-ability students by the instructional method

Dependent Variables

The Instructional Method CLMS CL T N= 40 N= 38 N= 42 Mean SD Adj. mean Std. Error 20.4500 1.6939 20.528 .244 17.8000 1.8701 17.968 .234

a a

Mathematics Performance (MP)

**20.000 1.2945 19.908 .250 16.5526 1.4275 16.374 .241
**

a a

**18.4286 1.6101 18.438 .236 14.6190 1.4808 14.620 .227
**

a a

Mathematical reasoning (MR)

Mean SD Adj. mean Std. Error

146

Metacognitive Knowledge (MK)

Mean SD Adj. mean Std. Error

**2.4317 .2244 2.419 .031
**

a

**2.2123 .1929 2.218 .032
**

a

**1.9492 .1861 1.956 .030
**

a

**Note. a. Evaluated at covariates appeared in the model: pre-MP = 11.3083, preMR = 7.8083.
**

Total score on MP = 22, total score on MR = 22, and total score on MK = 05

To examine if there were statistically significant differences in mathematics

performance, mathematical reasoning, and metacognitive knowledge adjusted mean scores between the high-ability students in CLMS group, in CL, and in T group, while controlling pre-MP and pre-MR, multivariate analysis of covariance (MANCOVA) was conducted.

Table 4.7 presents the results of multivariate analysis of covariance (MANCOVA), showing overall differences for the independent variable of instructional method effect on high-ability students and the three dependent variables, while controlling pre-MP and pre-MR. The Pillai’s Trace was used to evaluate the multivariate (MANCOVA) differences. The MANCOVA results of comparing the high-ability students across the three groups were statistically significant (F= 46.575, p = .000). The covariates preMP (F = 15.020, p = .002) and pre-MR (F = 16.553, p = .000) had significant effects. This means that there were some statistical differences between high-ability students across the three groups on at least one dependent variable. Further, the results of the univariate ANCOVA tests, which are represented in table 4.7, indicated that there were statistically significant differences between high-ability students across the three groups in the three dependent variables (MP, MR, and MK).

147

The F ratio of MP (2, 117) was 45.600 ( p = .000). This means that the instructional method had a main effect on high-ability students’ MP. This effect accounted for 26% of the variance of the high-ability students’ MP (Eta 2 = .258). The F ratio of MR (2, 117) was 162.490 ( p = .000). This means that the instructional method had a main effect on high-ability students’ MR. This effect accounted for 48% of the variance of the high-ability students’ MR (Eta2 = .477). The F ratio of MK (2, 117) was 202.729 (p = .000). This means that the instructional method had a main effect on high-ability students’ MK. This effect accounted for 49% of the variance of the high-ability students’ MK (Eta2 = .494).

Table 4.7 Summary of multivariate analysis of covariance (MANCOVA) results by the instructional method and follow-up analysis of covariance (ANCOVA) results of comparing high-ability students across the three groups.

MANCOVA Effect, Dependent Variables, and Covariate Group Effect Mathematics Performance(MP) Mathematical Reasoning(MR) Metacognitive knowledge (MK) Pre-MP Pre-MR 15.020 ( p = .002) 16.553 ( p = .000) Multivariate F Pillai's Trace 46.575 ( p = .000) 45.600 ( p = .000) 162.490 ( p = .000) 202.729 ( p = .000) Univariate F df = 2, 117

148

The MANCOVA results of comparing high-ability students across the three groups on the three dependent variables indicated that there were statistically significant differences between high-ability students in at least two groups on the three dependent variables. Therefore, the researcher further investigated the univariate statistics results (analysis of covariance ANCOVA) by performing a post hoc pairwise comparison using the /lmatrix command for each dependent variable in order to identify significantly where the differences in the adjusted means resided. Table 4.8 is a summary of post hoc pairwise comparisons between high-ability students across the three groups.

Table 4.8 Summary of post hoc pairwise comparisons between high-ability students across the three groups

Dependent Variable

Mathematics Performance (MP) Comparison Group CLMS vs. Adj.Mean Difference .620 Sig .081 Mathematical Reasoning (MR) Adj.Mean Difference 1.594 Sig .000 Metacognitive Knowledge (MK) Adj.Mean Difference .201 Sig .000

149

CL CLMS vs. T CL vs. T 2.090 .000 3.348 .000 .463 .000

1.471

.000

1.754

.000

.261

.000

Note.

The adjusted mean differences shown in this table are the subtraction of the second condition (on the lower line) from the first condition (on the upper line); for example, .620 (Adjusted Mean Difference for Mathematics Performance) = CLMS – CL.

Table 4.6 displays the means, standard deviations, adjusted means and standard errors of high-ability students in the three groups by the dependent variables. Table 4.7 and table 4.8 show that there are statistical adjusted mean differences among the highability students in the three conditions on the three dependent variables unless no statistical adjusted mean differences between high-ability students in CLMS and CL groups in mathematics performance. The adjusted mean differences are presented below.

Mathematics performance. The CLMS (Mean = 20.5, SD = 1.7, Adj.mean = 20.5) high-ability students and the CL (Mean = 20.0, SD = 1.3, Adj.mean = 19.9) highability students significantly outperformed the T high-ability students (Mean = 18.4, SD = .1.6, Adj.mean = 18.4) (p = .000), with adjusted mean differences of 2.090 and 1.471 respectively. There were no statistically significant differences between highability students in CLMS group and high-ability students in CL group (p = .081), with an adjusted mean difference of .620. (Effect sizes on MP were .28 and .98 for

150

comparing high-ability students in the CLMS and CL, and CL and the T group, respectively). Mathematical reasoning. The CLMS (Mean = 17.8, SD = 1.9, Adj.mean = 17.9) highability students significantly outperformed the CL and T high-ability students, with an adjusted mean difference of 1.594 (p = .000) and 3.348 (p = .000) respectively. The CL (Mean = 16.6, SD = 1.4, Adj.mean = 16.4) high-ability students significantly outperformed the T high-ability students (Mean = 14.6, SD = 1.5, Adj.mean = 14.6) with an adjusted mean difference of 1.754 (p = .000). (Effect sizes on MR were .84 and 1.3 for comparing high-ability students in the CLMS and CL, and CL and the T group, respectively). Metacognitive knowledge. The CLMS (Mean = 2.4, SD = .2, Adj.mean = 2.4) highability students significantly outperformed the CL and T high-ability students, with adjusted mean differences of .201 (p = .000) and .463 (p = .000) respectively. The CL (Mean = 2.2, SD = .2, Adj.mean = 2.2) high-ability students significantly outperformed the T high-ability students (Mean = 1.9, SD = .2, Adj.mean = 1.9) with an adjusted mean difference of .261 (p = .000). (Effect sizes on MK were 1.2 and 1.4 for comparing high-ability students in the CLMS and CL, and CL and the T group, respectively). 4.3.4 Summary of Testing Hypothesis 2 (CLMSH > CLH > TH) The statistical results partially support the hypothesis, that is, “CLMS H > CLH” is confirmed in mathematical reasoning and metacognitive knowledge while in mathematics performance is not. High-ability students taught via CLMS instructional method performed significantly higher than high-ability students taught via CL

151

instructional method in mathematical reasoning and metacognitive knowledge but they did not perform significantly higher in mathematics performance. “CLMSH, CL H > TH” is confirmed. High-ability students taught via CLMS and high-ability students taught via CL instructional methods performed significantly higher than the highability students taught via T instructional method in mathematics performance, mathematical reasoning, and metacognitive knowledge. 4.3.5 Testing of Hypotheses 3 Low-ability students taught via CLMS instructional method will perform higher than Low-ability students taught via CL instructional method who, in turn, will perform higher than low-ability students taught via T instructional method in (a) mathematics performance (MP), (b) mathematical reasoning (MR) and (c) metacognitive knowledge. Table 4.9 presents overall means, standard deviations, adjusted means, and standard error of each dependent variable for low-ability students by the instructional method, CLMS, CL, and T.

Table 4.9 Means, standard deviations, adjusted means and standard errors for each dependent variable for low-ability students by the instructional method

Dependent Variables

The Instructional Method CLMS CL T N= 40 N= 41 N= 39

152

Mathematics Performance (MP)

Mean SD Adj. mean Std. Error

**16.8500 1.2517 16.923 .198 14.5000 1.2195 14.662 .177 2.1633 .2086 2.170 .026
**

a a a

**15.2927 1.4533 15.367 .194 11.9512 1.4992 11.967 .174 1.7041 .1578 1.706 .025
**

a a a

**14.9744 1.3858 14.821 .199 10.6923 1.1955 10.509 .179 1.4821 .1008 1.472 .026
**

a a a

Mathematical reasoning (MR)

Mean SD Adj. mean Std. Error

Metacognitive Knowledge (MK)

Mean SD Adj. mean Std. Error

**Note. a. Evaluated at covariates appeared in the model: pre-MP = 8.9750, preMR = 3.2417.
**

Total score on MP = 22, total score on MR = 22, and total score on MK = 05

To examine if there were statistically significant differences in mathematics

performance, mathematical reasoning, and metacognitive knowledge adjusted mean scores between the low-ability students in CLMS group, in CL group, and in T group, while controlling pre-MP and pre-MR, multivariate analysis of covariance (MANCOVA) was conducted. Table 4.10 presents the results of multivariate analysis of covariance (MANCOVA), showing overall differences for the independent variable of instructional method effect on low-ability students and the three dependent variables, while controlling pre-MP and pre-MR. The Pillai’s Trace was used to evaluate the multivariate (MANCOVA) differences. The MANCOVA results of comparing the low-ability students across the

153

three groups were statistically significant (F = 27.918, p = .000). The covariates preMP (F = 12.202, p = .000) and pre-MR (F = 10.620, p = .000) had significant effects. This means that there were some statistical differences between low-ability students across the three groups on at least one dependent variable. Further, the results of the univariate ANCOVA tests, which are represented in table 4.10, indicated that there were statistically significant differences between low-ability students across the three groups in the three dependent variables (MP, MR, and MK). The F ratio of MP (2, 117) was 29.823 ( p = .000). This means that the instructional method had a main effect on low-ability students’ MP. This effect accounted for 34% of the variance of the low-ability students’ MP (Eta2 = .342). The F ratio of MR (2, 117) was 138.065 ( p = .000). This means that the instructional method had a main effect on low-ability students’ MR. this effect accounted for 71% of the variance of the low-ability students’ MR (Eta2 = .706). The F ratio of MK (2, 117) was 188.719 (p = .000). This means that the instructional method had a main effect on low-ability students’ MK. This effect accounted for 77% of the variance of the low-ability students’ MK (Eta2 = .766).

Table 4.10 Summary of multivariate analysis of covariance (MANCOVA) results by the instructional method and follow-up analysis of variance (ANOVA) results of comparing low-ability students across the three groups.

MANCOVA Effect, Dependent Variables, and Covariate Multivariate F Pillai's Trace Univariate F df = 2, 117

154

Group Effect Mathematics Performance(MP) Mathematical Reasoning(MR) Metacognitive knowledge (MK) Pre-MP Pre-MR

27.918 ( p =.000) 29.823 ( p = .000) 138.065 ( p = .000) 188.719 ( p = .000) 12.202 ( p = .000) 10.620 ( p = .000)

The MANCOVA results of comparing low-ability students across the three groups on the three dependent variables indicated that there were statistically significant differences between low-ability students in at least two groups on the three dependent variables. Therefore, the researcher further investigated the univariate statistics results (analysis of covariance ANCOVA) by performing a post hoc pairwise comparison using the /lmatrix command for each dependent variable in order to identify significantly where the differences in the adjusted means resided. Table 4.11 is a summary of post hoc pairwise comparisons between low-ability students across the three groups.

Table 4.11 Summary of post hoc pairwise comparisons between low-ability students across the three groups

Dependent Variable

Mathematics Performance Mathematical Reasoning Metacognitive Knowledge

155

(MP) Comparison Group CLMS vs. CL CLMS vs. T CL vs. T Adj.Mean Difference 1.555 Sig .000

(MR) Adj.Mean Difference 2.696 Sig .000

(MK) Adj.Mean Difference .464 Sig .000

2.101

.000

4.153

.000

.698

.000

.546

.053

1.458

.000

.234

.000

Note.

The adjusted mean difference shown in this table is the subtraction of the second condition (on the lower line) from the first condition (on the upper line); for example, 1.555 (Adjusted Mean Difference for Mathematics Performance) = CLMS – CL.

Table 4.9 displays the means, standard deviations, adjusted means and standard errors of low-ability students in the three groups by the dependent variables. Table 4.10 and table 4.11 show that there are statistical adjusted mean differences among the lowability students in the three conditions on the three dependent variables unless no statistical adjusted mean differences between low-ability students in CL and T group in mathematics performance. The adjusted mean differences are presented below.

Mathematics performance. The CLMS (Mean = 16.8, SD = 1.3, Adj.mean = 16.9) low-ability students significantly outperformed the CL (Mean = 15. 3, SD = 1.5, Adj.mean = 15.4) and the T (Mean = 14.9, SD = 1.4, Adj.mean = 14.8) low-ability students with adjusted mean differences of 1.555 (p = .000) and 2.101 (p = .000) respectively. However, there were no significant differences between low-ability

156

students in CL group and low-ability students in T group (p = .053), with an adjusted mean difference of .546. (Effect sizes on MP were 1.12 and .23 for comparing lowability students in the CLMS and CL, and CL and the T group, respectively). Mathematical reasoning. The CLMS (Mean = 14.5, SD = 1.2, Adj.mean = 14.7) lowability students significantly outperformed the CL and T high-ability students, with adjusted mean differences of 2.696 (p = .000) and 4.153 (p = .000) respectively. The CL (Mean = 11.9, SD = 1.5, Adj.mean = 11.9) low-ability students significantly outperformed the T low-ability students (Mean = 10.7, SD = 1.2, Adj.mean = 10.5) with an adjusted mean difference of 1.458 (p = .000). (Effect sizes on MR were 2.13 and 1.05 for comparing low-ability students in the CLMS and CL, and CL and the T group, respectively). Metacognitive knowledge. The CLMS (Mean = 2.7, SD = .2, Adj.mean = 2.2) lowability students significantly outperformed the CL and T high-ability students, with adjusted mean differences of .464 (p = .000) and .698 (p = .000) respectively. The CL (Mean = 1.7, SD = .2, Adj.mean = 1.7) low-ability students significantly outperformed the T low-ability students (Mean = 1.5, SD = .1, Adj.mean = 1.5) with an adjusted mean difference of .234 (p = .000). (Effect sizes on MK were 4.6 and 2.2 for comparing low-ability students in the CLMS and CL, and CL and the T group, respectively). 4.3.6 Summary of Testing Hypothesis 3 (CLMSL > CLL > TL) The statistical results partially support the hypothesis, that is, “CLMSL > CLL” and “CLMSL > TL” are confirmed in mathematics performance, mathematical reasoning, and metacognitive knowledge. “CLL > TL” is confirmed in mathematical reasoning and metacognitive knowledge while in mathematics performance is not. Low-ability

157

students taught via CLMS instructional method performed significantly higher than low-ability students taught via CL instructional method and than low-ability students taught via T instructional method in mathematics performance, mathematical reasoning, and metacognitive knowledge. Low-ability students taught via CL instructional method performed significantly higher than low-ability students taught via T instructional method in mathematical reasoning, and metacognitive knowledge but they did not perform significantly higher in mathematics performance. 4.3.7 Testing of Hypotheses 4 There are interaction effects between the instructional methods and the ability levels (high-ability and low-ability) on mathematics performance, knowledge. Table 4.12 presents overall means, standard deviations, adjusted means, and standard error of each dependent variable by the interaction between the instructional methods and the ability levels (high-ability and low-ability). mathematical reasoning, and metacognitive

Table 4.12 Means, standard deviations, adjusted means and standard errors for each dependent variable by the interaction between the instructional methods and the ability levels (high-ability and low-ability)

Dependent Variables Mathematic Mathematical s Reasoning Performance MR MP Metacognitiv e Knowledge MK

158

Instructional Method CLMS

Ability High (H) Low (L) H (n = 40) Mean SD Adj. mean Std. Error Mean SD Adj. mean Std. Error Mean SD Adj. mean Std. Error Mean SD Adj. mean Std. Error Mean SD Adj. mean Std. Error Mean SD Adj. mean Std. Error 20.4500 1.6939 19.681 .266 16.8500 1.2517 17.802 .285 20.000 1.2945 18.997 .296 16.8500 1.2517 16.226 .287 18.4286 1.6101 17.545 .279 14.9744 1.3858 15.734 .269 17.8000 1.8701 16.745 .249 14.5000 1.2195 15.833 .267 16.5526 1.4275 15.155 .277 11.9512 1.4992 13.213 .269 14.6190 1.4808 13.430 .262 10.6923 1.1955 11.723 .252 2.4317 .2244 2.440 .035 2.1633 .2086 2.159 .038 2.2123 .1929 2.219 .039 1.7041 .1578 1.690 .038 1.9492 .1861 1.964 .037 1.4821 .1008 1.472 .035

a

a

a

L (n = 40)

a

a

a

CL

H (n = 38)

a

a

a

L (n = 41)

a

a

a

T

H (n = 42)

a

a

a

L (n = 39)

a

a

a

**Note. a Evaluated at covariates appeared in the model: pre-MP = 10.1417, preMR = 5.5250.
**

Total score on MP = 22, total score on MR = 22, and total score on MK = 05

To examine if the effects of instructional method on mathematics performance, mathematical reasoning, and metacognitive knowledge depend on the ability level in

159

CLMS group, in CL, and in T group, while controlling pre-MP and pre-MR, two-way multivariate analysis of covariance (MANCOVA) was conducted. Table 4.13 presents the results of two-way multivariate analysis of covariance (MANCOVA), showing overall differences for the interaction between instructional method and ability level effect on the three dependent variables, while controlling preMP and pre-MR. The Pillai’s Trace was used to evaluate the multivariate (MANCOVA) differences. The MANCOVA results of the interaction effects on the three dependent variables was statistically significant (F = 4.836, p = .000). The covariates pre-MP (F = 15.020, p = .000) and pre-MR (F = 16.553, p = .000) had significant effects. This means that there were some statistical interaction effects on at least one dependent variable across the three groups. Further, the results of the two-way univariate ANCOVA tests, which are represented in table 4.13, indicated that there were statistically significant interaction effects across the three groups in MR and MK. The F ratio of MR (2, 237) was 3.401 (p=

.035). This means that the interaction effect was statistically significant on students’ MR. This interaction accounted for 3% of the variance of the students’ MR (Eta2 = .028). The F ratio of MK (2, 237) was 10.557 (p = .000). This means that the interaction effect was statistically significant on students’ MK. This interaction accounted for 8% of the variance of the students’ MK (Eta2 = .083). However, there were no statistically significant interaction effects across the three groups in MP. The F ratio of MP (2, 237) was 2.917 ( p > .05).

Table 4.13 Summary of multivariate analysis of covariance (MANCOVA) results by the interaction effect and follow-up analysis of covariance (ANCOVA) results across the three groups.

160

MANCOVA Effect, Dependent Variables, and Covariates Group Effect Mathematics Performance(MP) Mathematical Reasoning(MR) Metacognitive knowledge (MK) Pre-MP Pre-MR

Multivariate F Pillai's Trace 4.836( p = .000)

Univariate F df = 2, 237

2.917 ( p = .056) 3.401 ( p = .035) 10.557 ( p = .000) 15.020( p = .000) 16.553 ( p = .000)

The two-way MANCOVA results of the interaction effects on MR and MK indicated that there were statistically significant interaction effects between the instructional method and the students’ ability level in at least one group. Therefore, the researcher further investigated the interaction effect results by plotting the interaction between the instructional method and the students’ ability level on MR and MK to identify significantly where the interactions resided. Also the interaction between the instructional method and the students’ ability level on MP is plotted. Figure 4.1 shows the interaction effect between the instructional method and the students’ ability level across the three groups on MP.

161

20

19

MP Adjusted Mean Scores

18

17

Ability

16

High-ability

15 CLMS

Low-ability

CL T

Instructional Method

Figure 4.1 Interaction effect between the instructional method and the students’ ability levels on MP Figure 4.1 shows that there is no interaction effect between the instructional method and the students’ ability level on MP across the three groups. In other words, highability and low-ability students taught via CLMS, CL, and T instructional methods benefited equally in mathematics performance. Therefore, the effect of the instructional methods on MP did not depend on the ability level. Figure 4.2 shows the interaction between the instructional method and the students’ ability level across the three groups on mathematical reasoning (MR).

162

18

17

MR Adjusted Mean Scores

16

15

14

13

Ability

High-ability Low-ability

CL T

12 11 CLMS

Instructional Method

Figure 4.2 Interaction effect between the instructional method and the students’ ability levels on MR Figure 4.2 shows that the low-ability students taught via CLMS instructional method benefited more than the high-ability students taught via the same instructional method in mathematical reasoning. However, the figure shows that the high-ability and lowability students taught via CL and T instructional methods benefited equally in mathematical reasoning. Figure 4.3 shows the interaction between the instructional method and the students’ ability level across the three groups on metacognitive knowledge (MK).

163

2.6

2.4

MK Adjusted Mean Scores

2.2

2.0

1.8

Ability

1.6

**High-ability Low-ability
**

CL T

1.4 CLMS

Instructional Method

Figure 4.3 Interaction effect between the instructional method and the students’ ability levels on MK Figure 4.3 shows that the low-ability students taught via CLMS instructional method benefited more than the high-ability students taught via the same instructional method in metacognitive knowledge. However, the figure shows that the high-ability and lowability students taught via CL and T instructional methods benefited equally in metacognitive knowledge. 4.3.8 Summary of Testing Hypotheses 4 (There are interaction effects between the instructional methods and the ability levels) The statistical interaction results and the interaction figures partially confirm the hypotheses, showing that there were interaction effects between the CLMS instructional method and the ability levels where low-ability students benefited more than the high-ability students in MR and MK but benefited equally in MP. There were

164

no interaction effects between the CL instructional method and the ability level. That is, the performance of the CL instructional method did not depend on the ability level. High-ability and low-ability students taught via the CL instructional method benefited equally in MP, MR, and MK. Finally, there were no interaction effects between the T instructional method and the ability levels. That is, the performance of the T instructional method did not depend on the ability levels. High-ability and low-ability students taught via the T instructional method benefited equally in MP, MR, and MK.

4.3.9 Summary of Findings to Research Questions 1 – 4 The findings to the four research questions are summarized below. 1. Would students taught via CLMS instructional method perform higher than students taught via CL instructional method who, in turn, would perform higher than students taught via T instructional method in (a) mathematics performance (MP), (b) mathematical reasoning (MR) and (c) metacognitive knowledge (MK)? Overall, CLMS instructional method has significant positive effects on students’ (a) mathematics performance, (b) mathematical reasoning, and (c) metacognitive knowledge. This is evidenced by the statistical results that the students taught via the CLMS method significantly performed higher than the students taught via the CL and the students taught via the T methods in (a) mathematics performance, (b) mathematical reasoning, and (c) metacognitive knowledge. In addition, CL instructional method has significant positive effects on students’ (a) mathematics performance, (b) mathematical reasoning, and (c) metacognitive knowledge. This is evidenced by the statistical results that the students taught via the CL method significantly performed higher that the students taught via the T method in (a)

165

mathematics performance, (b) mathematical reasoning, and (c) metacognitive knowledge. 2. Would high-ability students taught via CLMS instructional method perform higher than high-ability students taught via CL instructional method who, in turn, would perform higher than high-ability students taught via T instructional method in (a) mathematics performance (MP), (b) mathematical reasoning (MR) and (c) metacognitive knowledge (MK)? CLMS instructional method has positive effects on high-ability students’ (a) mathematics performance, (b) mathematical reasoning, and (c) metacognitive knowledge. The high-ability students taught via the CLMS method significantly performed higher than the high-ability students taught via the T method in (a) mathematics performance, (b) mathematical reasoning and (c) metacognitive knowledge. In addition, except in (a) mathematics performance, the high-ability students taught via the CLMS method significantly performed higher than the highability students taught via the CL method. Also CL instructional method has significant positive effects on high-ability students’ (a) mathematics performance, (b) mathematical reasoning, and (c) metacognitive knowledge. The high-ability students taught via the CL method significantly performed higher than the high-ability students taught via the T method in (a) mathematics performance, (b) mathematical reasoning and (c) metacognitive knowledge. 3. Would low-ability students taught via CLMS instructional method perform higher than low-ability students taught via CL instructional method who, in turn, would perform higher than low-ability students taught via T instructional method in (a)

166

mathematics performance (MP), (b) mathematical reasoning (MR) and (c) metacognitive knowledge (MK)? CLMS instructional method has significant positive effects on low-ability students’ (a) mathematics performance, (b) mathematical reasoning, and (c) metacognitive knowledge. The low-ability students taught via the CLMS method significantly performed higher than the low-ability students taught via the CL and the T methods in (a) mathematics performance, (b) mathematical reasoning and (c) metacognitive knowledge. In addition, CL instructional method has significant positive effects on low-ability students’ (b) mathematical reasoning and (c) metacognitive knowledge. The low-ability students taught via the CL method significantly performed higher than the low-ability students taught via the T method in (b) mathematical reasoning and (c) metacognitive knowledge, but they did not perform significantly higher in (a) mathematics performance. 4. Are there interaction effects between the instructional methods and the ability levels (high-ability and low-ability) on mathematics performance,

mathematical reasoning, and metacognitive knowledge?

There were interaction effects between the CLMS instructional method and the ability levels with low-ability students benefited more than the high-ability students in mathematical reasoning (MR) and metacognitive knowledge (MK) but benefited equally in mathematics performance (MP). There were no interaction effects between the CL instructional method and the ability levels i.e., high-ability and low-ability students taught via the CL instructional method benefited equally in MP, MR, and MK. Finally, there were no interaction effects between the T instructional method and

167

the ability levels i.e., high-ability and low-ability students taught via the T instructional method benefited equally in MP, MR, and MK.

168

CHAPTER FIVE

DISCUSSION AND CONCLUSIONS 5.1 Introduction The purpose of this study was to investigate the effects of cooperative learning with metacognitive scaffolding (CLMS) and cooperative learning (CL) on (a) mathematics performance, (b) mathematical reasoning, and (c) metacognitive knowledge among fifth-grade students in Jordan. The study further investigated the effects of CLMS and CL on high-ability and low-ability students’ (a) mathematics performance, (b) mathematical reasoning, and (c) metacognitive knowledge. Students’ mathematics performance, mathematical reasoning, and metacognitive knowledge were measured through a mathematics achievement test and a metacognitive knowledge questionnaire. The sample consisted of 240 Jordanian male students who studied in six fifth-grade classrooms and were randomly selected from three different male primary schools i.e., two classes from each school. They studied “Adding and Subtracting Fractions” unit. The independent variable was the instructional method with three categories: Cooperative learning with metacognitive scaffolding instructional method (CLMS), Cooperative learning instructional method (CL), and Traditional instructional method (T). The moderator variable was the ability level with two categories: High-ability and Low-ability. The dependent variables were: Mathematics performance (MP), Mathematical reasoning (MR), and Metacognitive knowledge (MK). Data was collected during the first semester of the academic year 2002 / 2003. Two months before the instructional treatment, the participating students were given the

169

mathematics achievement test (pre-test). Students were randomly assigned to one of the three conditions – CLMS method, CL method, or T method. Then students were divided into high and low-abilities based on their pre-test scores in mathematics performance and mathematical reasoning. In CLMS method, students worked cooperatively and used metacognitive questions cards, while in CL method; students worked cooperatively and did not use metacognitive questions cards. In T method, students neither worked cooperatively nor used metacognitive questions cards. Immediately, after the instructional treatment, the students were given the mathematics achievement test (post-test) and the metacognitive knowledge questionnaire. In this chapter, interpretations of the results are discussed. The present chapter is organized in seven main sections. The first section focuses on the general effects of the instructional methods on mathematics performance, mathematical reasoning, and metacognitive knowledge. In the second section, the effects of the instructional methods on mathematics performance, mathematical reasoning, and metacognitive knowledge based on ability levels are discussed. The third section focuses on the interaction effects. The fourth section presents the summary and conclusions. The fifth section suggests implications for educators. The sixth section proposes implications for future research. Finally, the seventh section summarizes the limitations of the present study.

170

5.2 Effects of the Instructional Methods on Mathematics Performance, Mathematical Reasoning, and Metacognitive Knowledge CLMS and CL instructional methods had significant positive effects overall on students’ (a) mathematics performance, (b) mathematical reasoning, and (c) metacognitive knowledge. Students taught via the CLMS method (working cooperatively and also using metacognitive questions cards) significantly outperformed their counterparts taught via the CL method who, in turn, significantly outperformed the students taught via the T method in (a) mathematics performance, (b) mathematical reasoning, and (c) metacognitive knowledge. The findings on cooperative learning with metacognitive scaffolding (CLMS) support the hypothesis that cooperative learning with metacognitive scaffolding not only improves mathematics performance, as shown by the studies of Schoenfeld (1985); Peterson et al.(1982); Peterson et al. (1984); and King (1991a), but also improves mathematical reasoning and metacognitive knowledge. 5.2.1 Effects of the Instructional Methods on Mathematics Performance The effectiveness of CLMS method on mathematics performance that consists of conceptual understanding and procedural fluency support King and Rosenshine’s (1993) study that found that guidance through questioning enhances problem representation and improves conceptual understanding. The metacognitive questions have provided the students with cues to important aspects of the problem and helped them to identify the problem and identify relevant and important information. While conceptual understanding is enhanced by constructing relationships between the

171

previous and the new knowledge (Kilpatrick et al, 2001), the CLMS method encouraged students to identify the similarities and differences between the problem at hand and the problems solved in the past. The findings of this study are consistent with studies by Schonfeld (1987) and Xun (2001) that questioning strategies enabled students to connect what they learned with their current learning situation. Metacognitive questions helped students to make connections between different factors and constraints and link to the solutions. In this regard, metacognitive questions assisted students to enhance their understanding of a given domain knowledge. Flexibility, accuracy, and efficiency are fundamental components of procedural fluency (Kilpatrick et al, 2001). Students taught via the CLMS method were provided with the opportunity to execute their mathematical procedures fluently. Working cooperatively and using the metacognitive questions provided the students with more than one approach to solve the problem. Metacognitive question such as “what is the appropriate approach to …..?” helped the students to select the appropriate approach from many approaches to solve the problem. Because students asked questions such as “am I on the right track?, they were able to keep track of sub-problems and make use of intermediate results to solve the problem and therefore to be more accurate and more efficient learners. Thus, cooperative learning with metacognitive scaffolding

method enabled students to modify and adapt procedures to make them easier to use.

The high mathematics performance requires acquiring relevant conceptual understanding and procedural knowledge. The performance of the problem solver acts on these requisites. This probably accounts for the better performance of the students taught via CLMS method over the students taught via CL method who, in turn,

172

**performed better than the students taught via T method. Cooperative learning with
**

metacognitive scaffolding method enabled students to acquire the appropriate

procedural problem solving techniques, and therefore, they were able to maneuver the computations more accurately than the students in the other two groups. According to Cross and Parts (1988), this is the self-management aspect of metacognitive strategies. Students taught via the CL method worked cooperatively. The cooperative group provides a more intimate setting that permits such direct and unmediated communication (Shachar and Sharan, 1994). Such a context, proponents of cooperative learning believe, is a key to students engaging in real discussion and wrestling with ideas. Therefore, the CL method provided the students with the opportunities to stretch and extend their thinking more than the students taught via the T method who worked individually. The low performance of the students taught via the T method in this study emerged from the poor conceptual understanding and procedural techniques employed in solving tasks and problems. In the last meeting with the control group’s teacher, the teacher reported that students in this group worked individually and did not use metacognitive questions, did not plan, monitor, or evaluate their solution procedures, and mentioned that the students also immediately started the computations when the questions were given to them. Also some of students were anxious as to the specific demands of the questions. The control group’s teacher was satisfied with the performance of his group. The teacher’s report shows that students in the traditional group had insufficient conceptual understanding and procedural fluency and did not sufficiently or elaborately engage in the planning, monitoring and evaluation phases in solving their problems.

173

5.2.2 Effects of the Instructional Methods on Mathematical Reasoning The results of this study indicate that cooperative learning with metacognitive

scaffolding method enabled students to reflect on the similarities and differences

between previous and new tasks, as well as to comprehend each problem before attempting a solution, and to consider the use of strategies that are appropriate for solving the problem. The learning processes produced by the CLMS method enhanced the students’ mathematical reasoning. The effectiveness of CLMS method on

mathematical reasoning support other findings by Chi et al. (1994); Mevarech and Kramarski, (in press); Slavin (1996); and Webb (1989) that show that cooperative

learning with metacognitive scaffolding is one of the best means for elaborating

information and for making connections. By understanding why and how a certain solution to a task and a problem has been reached, the students elaborated on the information gained from the metacognitive questions and learned from it. Also Kramarski et., al. (2001, 2002) found that working cooperatively and using metacognitive questions facilitated metacognitive knowledge, which, in turn, affected mathematical reasoning and students’ ability to transfer their knowledge to solve mathematical authentic tasks. According to constructivist theories, information is retained and understood through elaboration and construction of connections between prior knowledge and new knowledge (Wittrock, 1986). The ability of constructing networks of knowledge with the CLMS method was greater than with the CL method which, in turn, was greater than with the T method. These findings are similar to Cossey (1997) findings that indicate that the more often seventh and eighth graders are exposed to metacognitive

174

support such as pattern seeking, conjectures, and giving reasons for ideas, the greater are their gains on mathematical reasoning. The findings of this study support earlier findings (Hoek et al., 1999; Mevarech, 1999) that cooperative learning with metacognitive scaffolding is effective for developing problem-solving ability because it enables students to link quantitative knowledge and situational knowledge. When the two types of knowledge are joined, a mental representation is constructed that supports mathematical reasoning (Cecil and Roazzi, 1994). The process of solving tasks at a high level of cognitive complexity (e.g., mathematical reasoning problems) depends on the activation of metacognitive processes more than on solving tasks at a lower level of cognitive complexity (e.g., conceptual and procedural problems) because the former requires careful planning, monitoring, regulation, and evaluation (Stein et al., 1996). The cooperative learning

with metacognitive scaffolding method forced students to activate such processes, so

they could reason mathematically better than the students taught via the CL method that focused only on working cooperatively and the students’ interaction was not structured. Specifically, the use of metacognitive questions guided students to analyze the entire situation described in the task or in the problem and thereby did not only enhance their understanding, but also enabled them to replace their earlier inappropriate strategies with a new virtually errorless process which is an essential element of mathematical reasoning. Students taught via the CLMS method could reason mathematically because they were guided about the knowledge of when, where, and why to use the strategies for the problem-solving. The metacognitive questions comprise of planning, monitoring, and

175

evaluation questions. Students taught via the CLMS method were required to plan, monitor, and evaluate their learning strategies and solutions. Planning questions enabled students to formulate, identify, and to define the task or the problem and then build the relationships among its concepts and procedures. Monitoring questions enabled students to regulate or monitor their problem performance by self-generating feedback which enabled them to select the appropriate strategies. Evaluation questions enabled students to reflect on their solutions or alternatives so as to direct their future steps. One of the most important components of mathematical reasoning is the appropriate strategies selection and the justification of selecting these strategies. The students taught via the CLMS method, were able to select and justify the appropriate strategies for solving the problem because they were trained how to do so. They were trained to ask metacognitive questions such as “what is the appropriate strategy to solve …? And “how do we justify the appropriateness of our strategy?” Also mathematical reasoning requires applying strategies in other situations. The students taught via the CLMS method were supported to generalize their learning strategies to other situations. Questions such as “how do we apply this line of thinking to other situations?” and “can we derive a rule that would work for …?” enabled students to generalize their strategies, and therefore enabled them to reason mathematically more than the other two groups. According to Piaget (1970), students work with independence and equality on each other’s ideas. The students taught via the CLMS method encountered situations that contradicted their believes or understanding. This is what Piaget calls cognitive conflicts. This conflict created a case of disequilibrium for the students. Metacognitive

176

questions that comprise planning, monitoring, and evaluation questions assisted students to assimilate or accommodate their knowledge and therefore reequilibrate their thinking. The students taught via the CLMS method were forced to revise, evaluate, and guide their ways of thinking to provide a better fit with reality and therefore their ability to reason mathematically was improved. Since learning with understanding according to Piaget occurs by assimilation or accommodation through resolving the cognitive conflicts, the students taught via the CLMS method were actively able to adjust and construct their knowledge set and strategies to settle disputes and disagreements and then their knowledge was assimilated or accommodated. Also when the students discussed with each other, different point of views emerged which pushed cognitive development by causing disequilibrium, which directed students to rethink their ideas. This learning situation created cognitive conflicts between the students and within every student which helped then to reason mathematically to reequilibriate their thinking. For Vygotsky (1978), students can be scaffolded by explanation, demonstration, and can attain to higher levels of thinking. Vygotsky (1978) suggests the ZPD which is the difference between what students can accomplish independently and what they can achieve under support and guidance. The students taught via the CLMS method were provided with the opportunity to be able to attain higher levels of knowing which were facilitated by the interaction between the low-ability and the high-ability students. Working cooperatively and using metacognitive questions provided students with the opportunity to explain, modify, and justify their solutions which, in turn, enabled them to extend themselves to higher levels of mathematical reasoning. In other words, students taught via the CLMS method were scaffolded through the cooperation i.e.,

177

high-ability and low-ability interaction, and through the use of metacognitive questions which helped the students in narrowing their ZPD. The students taught via the CL method were scaffolded through only the cooperation which enabled them to reason mathematically better than the students taught via the T method whose learning were not scaffolded. Therefore, when the metacognitive scaffolding was provided to students for group cooperation, the benefits of cooperative learning in mathematical reasoning were maximized. While the findings of this study confirmed previous research (Lin et al., 1999; Palincsar et al., 1987; Webb, 1982, 1989b; Brown and Palincsar, 1989; Kramarski et al., 2001, 2002) on the effectiveness of cooperative learning in supporting students’ mathematical reasoning and cognitive and metacognitive development, they also suggest that there were certain conditions in which the use of cooperative learning fully worked to facilitate learning. Greene and Land (2000) found that cooperative learning was useful in influencing the development of ideas only when group members offered suggestions, when they were open to negotiation of ideas, and when they shared prior experiences. There may be times when group members do not know how to ask questions or how to elaborate thoughts, or there may be times when group members are not willing to ask questions or respond to others’ questions, or there may be times when group members do not see the need for cooperation. Webb’s (1989b) model of cooperative learning further revealed that different conditions and patterns of cooperation might lead to different learning outcomes. Webb (1989b) found that the students who learned most were those who provided explanations to others in their group. In this regard, metacognitive questions served to facilitate the cooperative learning processes through eliciting responses from some students, and the responses may invoke further questions from other students who may require elaboration,

178

reasoning, or explanation from their peers. In this study the cooperative learning of the students taught via the CLMS method was structured and guided by the metacognitive questions cards and therefore these students were assisted to explain and reason their solution processes. 5.2.3 Effects of the Instructional Methods on Metacognitive Knowledge The effectiveness of the cooperative learning with metacognitive scaffolding method on metacognitive knowledge confirms the results of previous studies (e.g., Lin and

Lehman, 1999; Davis and Linn, 2000; King, 1991a, 1991b; Palincsar and Brown, 1984, 1989), which were all consistent in concluding that cooperative learning and questioning strategies enhanced metacognitive knowledge and reflective thinking. The use of metacognitive questions directed students’ attention to plan, monitor, and evaluate their learning processes, which helped them to obtain metacognitive knowledge and transfer their understanding to novel problems and situations. Also cooperative learning with metacognitive scaffolding directed students’ attention to relevant information which made them aware of the important factors and aspects to be considered, which in turn helped them to monitor their own understanding. The findings of this study confirm that the cooperative learning with metacognitive

scaffolding method facilitated metacognitive thinking by directing the students’

attention. The evaluation questions in the metacognitive questions card, such as “what are the evidences to justify…?” helped the students to reflect upon and explain their own actions and decisions. The findings of this study support Chi et al.’s (1989) and Lin and Lehman (1999) findings that metacognitive questions and self-explanation facilitated problem-solving processes and assisted students to make arguments for

179

their solutions and decisions, and thus make thinking explicit. Metacognitive questions also helped students to monitor their status of understanding in their problem solving processes by constantly referring back to the goals of the problem. Masui and De Corte (1999) findings show that students who used metacognitive questions had more knowledge about orienting and self-judging themselves than did students in the control groups. The cooperative learning with metacognitive scaffolding method helped students to elicit responses and explanations, which promote comprehension of the one who received the explanation and the one who gave the explanation and feedback (Webb, 1989b). The cooperative learning with metacognitive scaffolding method guided students to develop solutions by building upon each other's ideas, questioning each other, providing feedback, and checking the solution process. The students also were forced to check each other’s ideas to test if the selected solution was feasible or not, which required justification for a solution or suggestion, which facilitated the continuous monitoring of the problem-solving process. The students taught via the CLMS method were provided with multiple perspectives. Multiple perspectives gave an opportunity for students to reflect upon and evaluate their solution processes as the findings of Lin et al. (1999) showed. The choices of perspectives direct students’ attention to the important aspects of the problem that they might not have thought about, and as a result, students re-examine their thinking process, elaborating or modifying their thoughts, recognizing limitations in their solutions, or making justifications for their solutions or decisions. In this regard, CLMS method facilitated students’ metacognitive knowledge in the problem-solving process through planning, monitoring and evaluation. Metacognitive questions

180

enhanced planning by activating prior knowledge and attending to important information, monitoring by actively engaging students in their learning process, and enhanced evaluation through reflective thinking. A possible reason that the students taught via the CLMS method outperformed their counterparts taught via the CL method who, in turn, outperformed the students taught via the T method in metacognitive knowledge is that the CLMS method forced students to ask more metacognitive questions than the CL method which, in turn, forced students to ask more thinking and hinting questions than the T method. Students taught via the CLMS method were constantly trained to produce metacognitive questions and responses. The production of these questions, responses, and feedback during the cooperative setting promoted higher level thinking and understanding, and thus more metacognitive knowledge for participating students. Previous research (Flavell, 1979; Palinscar and Brown, 1984) showed that learning strategies that used cooperative learning and questioning activities function as a testing mechanism that allows students to monitor their own comprehension. It also helps students to realize what they know and more importantly, it helps the students to know what they do not know (King, 1989). Therefore, the students taught via the CLMS, who utilized this cooperative questioning strategy more extensively than the CL and T students, reported higher metacognitive knowledge levels than the other two groups. The students taught via the CL method worked cooperatively where multiple responses were provided. This learning environment somewhat encouraged students to produce high level thinking questions and provide evidence for their solutions more than the T students.

181

5.3 Effects of the Instructional Methods on Mathematics Performance, Mathematical Reasoning, and Metacognitive Knowledge Based on Ability Levels The results of this study showed that the high-ability students taught via CLMS method (high-ability students working cooperatively and also using metacognitive questions cards) significantly outperformed their counterparts taught via the CL method in (b) mathematical reasoning and (c) metacognitive knowledge. The highability students taught via the CLMS method and the high-ability students taught via the CL method significantly outperformed their counterparts taught via the T method in (a) mathematics performance, (b) mathematical reasoning, and (c) metacognitive knowledge. However, there were no statistically significant differences between the high-ability students taught via the CLMS method and the high-ability students taught via the CL method in (a) mathematics performance. Also the results showed that the low-ability students taught via the CLMS method (low-ability students working cooperatively and also using metacognitive questions cards) significantly outperformed their counterparts taught via the CL method and taught via the T method in (a) mathematics performance, (b) mathematical reasoning, and (c) metacognitive knowledge. The low-ability students taught via the CL method significantly outperformed their counterparts taught via the T method in (b) mathematical reasoning and (c) metacognitive knowledge. However, there were no statistically significant differences between the low-ability students taught via the CL method and the low-ability students taught via the T method in (a) mathematics performance.

182

5.3.1 Performance of High-Ability Students Taught Via CLMS The high-ability students taught via the CLMS method outperformed the high-ability students taught via the CL method in (b) mathematical reasoning and (c) metacognitive knowledge, and outperformed the high-ability students taught via the T method in (a) mathematics performance, (b) mathematical reasoning, and (c) metacognitive knowledge. Basically, high-ability students have acquired a great deal of content knowledge that is organized in ways that reflect a deep understanding of their subject matter, and therefore the high-ability students habitually use active learning strategies needed to monitor understanding (Golinkoff, 1976; Meichenbaum, 1976; Ryan, 1981). In addition to their habitual use of learning strategies, the high-ability students taught via the CLMS method worked cooperatively and were provided with metacognitive questions which assisted them to discuss, explain, and evaluate their and other students’ learning processes. Also the CLMS method gave the opportunity to the highability students to direct the low-ability students’ attention to the relevant features of the problem they could not understand. Through directing and guiding the low-ability students, the high-ability students’ reasoning, argumentation and justification were supported. Working cooperatively and using metacognitive questions further gave the opportunity to the high-ability students to actively engage in negotiation and meaning sharing, they asked metacognitive questions that challenged one’s thinking and required planning, monitoring, explanations, elaboration, and evaluation and justifications. In such an environment, the CLMS method created a setting for the high-ability students to construct arguments, reason, and make justifications. The CLMS method forced the high-ability students to ask the low-ability students and

183

themselves questions before, during, and after the solution processes. Asking and receiving answers of these metacognitive questions assisted the high-ability students to analyze the whole situation described in the problem, focus on the similarities and differences between previous and new tasks, as well as on comprehending the problem before attempting a solution and reflecting on the use of strategies that are appropriate for solving the problem, and thus enhanced their understanding and enabled them to evaluate, justify, and alter the inappropriate strategies with a new virtually errorless process which are an essential elements of mathematical reasoning and metacognitive knowledge. Working cooperatively with the low-ability students and asking and answering metacognitive questions provided the high-ability students with multiple perspectives and guided them to see things they might have overlooked. Also formulating and answering metacognitive questions forced the high-ability students to identify the main ideas and the ways the ideas relate to each other and to the students’ prior knowledge and experiences. Such a characteristic assisted the high-ability students to reflect on their own thinking, actions, and decisions, and as a result, they modified their thinking, planned remedial actions, evaluated their solutions, and monitored and checked their and other students’ solution processes. Additionally, the CLMS method assisted the high-ability students’ mathematics performance that comprises conceptual understanding and procedural fluency. Metacognitive questions within the cooperative learning setting provided the highability students with prompts to important features of the task and helped them to recognize the problem and recognize relevant and important information. Also the CLMS method encouraged the high-ability students to perceive the similarities and

184

differences between the current and the problems already solved previously. Working cooperatively and asking and answering metacognitive questions assisted the highability students to create connections between different aspects and constraints and relate to the solutions. In this regard, the CLMS method improved the high-ability students’ conceptual understanding and mathematical procedures. However, the findings of this study showed that although the adjusted mean of the high ability students taught via the CLMS method was higher, there were no statistically significant differences between the high-ability students taught via the CLMS method and the high-ability students taught via the CL method in (a) mathematics performance. Thus, unlike mathematical reasoning and metacognitive knowledge, metacognitive scaffolding did not assist the high-ability students to outperform their counterparts taught via the CL method in mathematics performance. This is due to the nature of the tasks and the problems that required conceptual understanding and procedural fluency, the two major elements of mathematics performance which were within their mastery. The processes of solving these tasks and problems require mastering the procedures and applying these procedures step by step more than the activation of metacognitive strategies which, mathematical reasoning and metacognitive knowledge tasks and problems require. Additionally, the high-ability students habitually often use active learning strategies needed to monitor understanding (Golinkoff, 1976; Meichenbaum, 1976; Ryan, 1981). The high-ability students taught via the CLMS method worked cooperatively with the low-ability students and asked and answered metacognitive questions. In this situation, the CLMS method guided the high-ability students to establish learning goals for tasks and problems, to assess the degree to which these

185

goals are being met, and, if necessary, to modify the strategies being used to meet the goals. Also the CLMS method forced the high-ability students to discuss with and ask the low-ability students metacognitive questions before, during, and after the processes of solving the mathematical tasks and problems. Therefore, the CLMS method assisted and guided the high-ability students to activate their metacognitive processes, and aided them to focus on formulating and understanding the problem more than on mastering the procedures of solving the problem. Working cooperatively and asking and answering metacognitive questions before, during, and after the processes of solving the problem assisted the high-ability students to focus on the processes of solving problems at a higher level of cognitive complexity, and thus, they were more guided to execute the tasks and problems that required mathematical reasoning and metacognitive knowledge than those that required conceptual understanding and procedural fluency. The effect of this activity is evidenced by the higher attainment in mathematical reasoning and metacognitive knowledge i.e., the high-ability students taught via the CLMS method outperformed their counterparts taught via the CL method in mathematical reasoning and metacognitive knowledge. 5.3.2 Performance of High-Ability Students Taught Via CL The findings of this study showed that the high-ability students taught via the CL method outperformed their counterparts taught via the T method in (a) mathematics performance, (b) mathematical reasoning, and (c) metacognitive knowledge. Working cooperatively with the low-ability students, the CL method gave an opportunity to the high-ability students to discuss, clarify ideas, and evaluate each others’ ideas. According to Vygotsky (1978), students are capable of performing at higher levels when working cooperatively than when working individually. Group diversity in

186

terms of knowledge and experience contributes positively to the learning process. Within the cooperative learning environment, the high-ability students are confronted with different interpretations of a given situation, and thus, the CL method created cognitive conflicts among the students which then enhanced them to discuss, explain, evaluate, and modify their opinions to reequilibriate their thinking to learn with understanding. Also the CL method provided the high-ability students with opportunities to learn from each other’s skills and experiences. Working cooperatively helped the high-ability students to go beyond simple statements of opinion by giving reasons for their judgments and reflecting upon the criteria employed in making these judgments. Thus, each opinion was subject to careful scrutiny. The ability to admit that one’s initial opinion may have been incorrect or partially flawed improved the high-ability students’ mathematical reasoning and metacognitive knowledge. 5.3.3 Performance of Low-Ability Students Taught Via CLMS The higher mathematics performance, mathematical reasoning, and metacognitive knowledge scores of the low-ability students taught via the CLMS method is explained by the fact that within cooperative setting, the metacognitive scaffolding guided the low-ability students in the right direction through the metacognitive questions and the questions generated by the high-ability students. Questions on planning, monitoring, and evaluation guided the low-ability students to construct sound arguments, evaluate solutions, and explain reasons for viable alternative solutions. This indicates that the low-ability students during problem solving need support and guidance in the problem solving process. The CLMS method forced each student to be asker, summarizer, recorder, and presenter by rotation. Working cooperatively with high-ability students and using the metacognitive questions

187

assisted the low-ability students to generate more questions which served as a guideline to help them start the problem solving task. The CLMS method gave the opportunity to every student to ask questions before, during, and after the processes of solving the problem, to generate more questions among the group members, and to elaborate thoughts in responding to those questions. The findings of this study overcome learning deficiencies involving low ability students as found by Graesser and Person (1994) and others. Graesser and Person (1994) found that low-ability students usually asked low frequency, short answer, and shallow questions and argue that this phenomenon can be attributed to difficulties at three different levels, one of them being the low-ability students’ difficulty identifying their own knowledge deficits, unless students had high amounts of domain knowledge. Gavelek and Raphael (1985) also pointed that low-ability students may lack the background knowledge necessary to ask their own questions or even answer the questions of others; and they may also lack the procedural knowledge for discriminating what it is that they do know from that which they do not know. Xun (2001) indicates that if the frequency of questions is low or if the questions asked are superficial, there would not be many explanations elicited from other students or even themselves. As found by Chi et al. (1989), self-explanation was an important component to monitor one’s learning process. In this study, the low-ability students taught via the CLMS method were provided with metacognitive questions and worked cooperatively, which assisted them to ask important and relevant questions before, during, and after the processes of solving the problem which, in turn, helped to elicit their responses, elaborate their thinking and articulate their reasoning and therefore, they solved the problems more correctly than

188

the other two groups. Whether the response is verbally articulated or thoughtfully considered, answering one’s own questions in the form of self-explanation can be an effective strategy for enhancing reflection and metacognition (Chi et al., 1989). Metacognitive scaffolding is especially important for low-ability students who tend to jump immediately into computation aspects of problem solving when faced with the task of solving complex problems (Lin et al., 1999). The low-ability students in this study may lack the ability to engage in effective thinking and problem solving on their own; thus the CLMS method enabled them to induce higher-order thinking and provided them with tools that they did not already possess. Also the CLMS method guided low-ability students’ attention to specific aspects of their learning process such as planning, monitoring and evaluation of their own problem-solving processes before, during and after the processes of solving the problems, which enhanced their metacognitive knowledge. 5.3.4 Performance of Low-Ability Students Taught Via CL The low-ability students taught via the CL method worked with the high-ability students together to solve problems and complete tasks. In this setting, the low-ability students had the opportunity to model the study skills and work habits of more proficient students. Through cooperation, the low-ability students were provided with different perspectives which helped them to evaluate and justify their solution processes and therefore they outperformed their counterparts who taught via the T method in (b) mathematical reasoning and (c) metacognitive knowledge. However, the findings of this study showed the although the mean of the low ability students in the CL group was higher there were the findings of this study showed that there were no statistically significant differences between the low-ability students taught via the CL

189

method and the low-ability students taught via the T method in (a) mathematics performance. In this study, mathematics performance tasks required conceptual understanding and procedural fluency. The process of solving tasks required conceptual understanding and procedural fluency does not depend on the activation of metacognitive processes as much as process of solving tasks required mathematical reasoning. The low-ability students taught via the CL method worked cooperatively with the high-ability students. This learning environment encouraged the low-ability students to discuss with and ask the high-ability students questions regarding the processes of solving the mathematical tasks and problems which, in turn enhanced the high-ability students to provide the low-ability students with multiple perspectives, guide them to activate their metacognitive processes, and assisted them to concentrate on formulating and understanding the problem more than on the procedures of solving the problem. The low-ability students taught via the CL method were assisted to focus on solving problems at a higher level of cognitive complexity, and therefore, they concentrated more on tasks and problems required mathematical reasoning and metacognitive knowledge than those required conceptual understanding and procedural fluency. This concentration is evidenced by outperforming their counterparts taught via the T method in mathematical reasoning and metacognitive knowledge. 5.3.5 Performance of Low-Ability Students Taught Via T The explanations for the low mathematics achievement and metacognitive knowledge of the low-ability students taught via the T method could be that they were not taught the appropriate strategies, could not self-regulate the study strategies, and did not understand how to apply these strategies. In the last meeting with the teacher who

190

applied the T method, the teacher reported that most of the low-ability students were confused when they encountered a mathematical problem and they were unable to explain the strategies they employed to find the correct solution. The teacher’s report confirmed that the low-ability students taught via the T method generally lack welldeveloped mathematical reasoning and metacognitive knowledge. The low-ability students taught via the T method were not scaffolded via cooperation and metacognitive questions, and thus they might have been at a loss as to how to start to solve the problems. They might not have known what questions to ask, they might not generate many questions to ask; or even if they did ask questions, the questions might not be focused or in-depth. Thus, they had limited abilities to solve problems require mathematical reasoning and metacognitive strategies. Within the traditional teaching method, the low-ability students often received less teacher time, attention, and were asked a fewer number of process-oriented questions (Leder, 1987). This may happened with the low-ability students taught via the T method because the teacher reported that he himself determines the success of the low-ability students, and thus may not gave these students more time and attention, and may not encouraged them to participate in the whole class public interaction. Also, the low-ability students taught via the T method were given much greater time and emphasis to mathematical procedures.

191

5.4 Interaction Effects An interesting finding in this study is that the low-ability students taught via the CLMS method benefited more than the high-ability students taught via the same method in (b) mathematical reasoning and (c) metacognitive knowledge. This finding is interpreted within the context of Piaget’s and Vygotsky’ theories. Seen according to Piaget’s cognitive-development theory (Piaget, 1970), the CLMS method played a critical role in enhancing cognitive development of low-ability students. The CLMS method created cognitive discrepancies or cognitive conflicts and therefore encouraged the students to resolve them. The causal sequence began with the metacognitive questions that generated tension while creating the discrepancy, which in turn caused disequilibrium, and the student then strived to resolve the discrepancies via mental activity. In this case, CLMS method challenged low-ability students to change their cognitive structure or schema to make sense of the environment, to think about alternative solutions and consider various perspectives. The CLMS method encouraged low-ability students’ mathematical reasoning and metacognitive knowledge to settle disputes and disagreements and then the knowledge was assimilated and accommodated. The metacognitive questions, particularly the “why” questions activated the low-ability students’ prior knowledge related to the new concepts. Therefore, the metacognitive questions helped to activate the low-ability students’ schemata and thus enabled them to retrieve information, elaborate knowledge, and represent understandings of the problem to be solved. It can be concluded that the low-ability students who are habitually deficient in the use of active learning strategies needed to monitor understanding (Golinkoff, 1976; Meichenbaum, 1976; Ryan, 1981) were through MS enhanced to perform like the high

192

ability students. The high ablity students were already habitual users of the processes of metacognition (Golinkoff, 1976; Meichenbaum, 1976; Ryan, 1981) and thus despite working cooperatively and asked and answered metacognitive questions did not benefit as much from the CLMS method. Low-ability students in the traditional teaching were rarely exposed to high-level reasoning and mathematical discussions (Mevarech and Kramarski, 1997), and frequently received less teacher time, attention, and were asked a fewer number of process-oriented questions (Leder, 1987). Therefore, the low-ability students were in a critical need of a learning method like the CLMS method that challenged them and then forced them to attend to the instructions sufficiently. The CLMS method assisted the low-ability students to organize the new material, integrate the information with existing knowledge and guide the encoding of schema. Also, the low-ability students taught via the CLMS method were supported to construct their mathematical knowledge by their own and therefore learn mathematics with understanding. Central to the notion of working cooperatively and using metacognitive questions is Vygotsky’s (1978) zone of proximal development, that is, “the distance between the actual development level as determined by independent problem solving and the level of potential development as determined through problem solving under adult guidance, or in collaboration with more capable peers” (p. 86). Through working cooperatively with high-ability students and asking and answering metacognitive questions, the low-ability students were provided with modeling of higher-level thinking and more sophisticated ways of constructing arguments, understanding textual materials, and solving problems, and thus the low-ability students reached levels in mathematics achievement and metacognitive knowledge that they could not

193

reach without cooperation and metacognitive scaffolding. Therefore, the CLMS method helped the low-ability students to fully utilize their potential abilities and to progress from what Vygotsky called their “actual developmental level” to their “level of potential development” (1978, p. 86). The high ability students in all groups were already independently functioning at ZPD levels which were higher than those of the low ability students. The CLMS method had a positive effect on the ZPD levels of high ability students but dramatically enhanced the ZPD levels the low ability students.

194

5.5 Summary and Conclusions This study found that the use of metacognitive scaffolding helped the students to fully benefit from cooperative learning. Overall the CLMS group outperformed the CL group in all measures, showing that for fifth-grade mathematics cooperative learning alone was not sufficient as a form of scaffolding. The low-ability students taught via the CLMS method outperformed their counterparts taught via the CL and T methods in mathematics performance, mathematical reasoning, and metacognitive knowledge. The low-ability students taught via the CL method in turn outperformed their counterparts taught via the T method in MR and MK but not in MP. This study shows that the cooperative learning method, when embedded with metacognitive scaffolding and implemented correctly in the

classrooms, is an effective method to achieve the goal of helping low-ability students learn mathematics with understanding, reason mathematically, and obtain and apply metacognitive strategies. The high-ability students taught via the CLMS method outperformed their counterparts taught via the CL method in MR and MK but not in MP, and outperformed their counterparts in the T method in MP, MR and MK. The high-ability students taught via the CL method in turn outperformed their counterparts taught via the T method in MP, MR and MK. The CLMS method was highly effective in the teaching of conceptual understanding and procedural fluency (mathematics performance) for both high-ability and lowability students, but the interaction effects showed that the CLMS method is very

195

effective for enhancing mathematical reasoning and metacognitive knowledge among low-ability students. From these findings, it can be concluded that the use of metacognitive scaffolding helped the students to fully benefit from cooperative learning. When students are actively engaged in activities such as planning, monitoring, questioning, explaining, elaborating, negotiating meanings, constructing arguments, and evaluation, they benefit much from the cooperative learning process. Therefore, the cooperative learning method is inadequate without metacognitive scaffolding or, cooperative learning with metacognitive scaffolding method is superior to cooperative learning method alone. It follows that the cooperative learning process should be scaffolded appropriately, and modeling through metacognitive scaffolding. The metacognitive scaffolding is especially effective in improving students’ mathematical reasoning and metacognitive knowledge. The cooperative learning with metacognitive scaffolding method is effective for younger students and for improving performance in all aspects of mathematics. The cooperative learning with metacognitive scaffolding method further is an effective method across abilities, but is especially beneficial for low ability students.

.

196

5.6 Implications for Educators From the discussion of the findings, it is evident in this study that cooperative learning with metacognitive scaffolding method is effective in supporting students’ mathematics performance, mathematical reasoning, and metacognitive knowledge. A close examination of the results revealed that cooperative learning alone is insufficient as a form of scaffolding. Also it is inferred that the metacognitive scaffolding was particularly effective in supporting low-ability students’ mathematics performance, mathematical reasoning, and metacognitive knowledge. Therefore, metacognitive scaffolding can be integrated in instructional design, curriculum design, computer based design, or web-based design to develop mathematics performance, mathematical reasoning, and metacognitive knowledge and facilitate self-regulated learning (Brown and Palincsar, 1989). The implementation of cooperative learning with metacognitive scaffolding method is not costly. Therefore, the effectiveness, the high learn ability level and the cost effectiveness of this method make this method a good candidate for inclusion in the development of the pedagogical approach. The cooperative learning with metacognitive scaffolding method should be included in teacher education programs. There are several skills, such as grouping, drawing metacognitive questions, and reflection, that pre-service and in-service teachers need to be trained. Also the use of cooperative learning with metacognitive scaffolding method in the classroom requires an approach to assessment and evaluation that is different from the present system. A more authentic and performance-based assessment criteria, that pre-service and in-service teachers need to be trained to develop to accompany the implementation of this method in the classroom.

197

In the usual manner, low-ability students do not get the same attention and do not have the knowledge and skills as high-ability students. In this study, the findings showed that the cooperative learning with metacognitive scaffolding is particularly effective in supporting low-ability students’ mathematics performance, mathematical reasoning, and metacognitive knowledge. Therefore, teachers should give low-ability students more attention and guide and assist them metacognitively. If low-ability students receive more attention and assisted metacognitively, they can perform almost as highability students as the findings of this study proved. Finally, at present, many state proficiency tests and international examinations (e.g., TIMSS- 1999 or PISA- 2000 administered by OECD- 2000 countries) include problems and tasks that ask students to explain their reasoning in writing. To acquaint students with such tasks, teachers should use metacognitive questions cards as guidelines and ask students to score one another’s reasoning by using the metacognitive questions cards and activating metacognitive processes.

198

5.7 Implications for Future Research The findings of this study raise several questions for further research: First, no formal observations and / or interviews were conducted in this study. Therefore, the quality of group interactions in CLMS and CL methods is not known. It would be particularly interesting to examine how high-ability and low-ability students interact with each other in the CLMS and the CL methods. Second, Students’ motivation is an interesting area for future research. Webb and Palincsar (1996) point out, “Groups are social systems. Students’ interaction with others is not only guided by the learning task, it is also shaped by their emotions, perceptions, and attitudes. Some social-emotional processes are beneficial for learning, others are not." (p. 855). Therefore, the effect of cooperative learning with metacognitive scaffolding and cooperative learning on students’ motivation is worth further investigation. Third, the findings of this study call for the design of additional learning environments based on similar components. The extent to which the CLMS and CL methods used in the present study are effective also for children at different grades, different gender, different mathematical topics, or for different subjects is not known at present and may be investigated in future research. Finally, an interesting question raised in this study relates to the effects of providing metacognitive scaffolding in cooperative learning setting versus providing metacognitive scaffolding in an individual learning setting on mathematics performance, mathematical reasoning, and metacognitive knowledge. To address the issue, students who worked cooperatively and used metacognitive questions cards

199

should be compared with students work individually and use metacognitive questions cards.

200

5.8 Limitations of the Study This study sought to investigate the effects of cooperative learning with metacognitive scaffolding and cooperative learning strategies on mathematics performance, mathematical reasoning, and metacognitive knowledge among fifth-grade students of two ability levels (high and low) in Jordan. This study was conducted in the natural setting of the class. The following are some limitations may restrict the probability of generalizing its findings: First, Jordan Government schools are not coeducational, so this study samples limited to the male fifth-grade students in the primary schools of Irbid directorate. The results found in this study may not be generalizable to the female fifth-grade students in other educational directorates in Jordan. Second, this study limited to the “Adding and Subtracting fractions” unit in the fifthgrade textbook, and this may restrict generalizing the study findings to the rest of mathematics concepts and subjects. The third limitation was associated with measuring students’ metacognitive knowledge. Students’ metacognitive knowledge was assessed via a metacognitive questionnaire; this assessment procedure may be insufficient as there are no verbal report measures or/and direct observation of students’ interaction and strategy use and development. However, the researcher justifies that the students’ age (11 years) may restrict the implementation of these assessment procedures. Therefore, the researcher relied primarily on the questionnaire data for the investigation of the students’ metacognitive knowledge.

201

REFERENCES Abu Libdeh, K. (2000, November). Analysis of Students' Errors in Mathematics Problem Solving. Paper presented at the international conference on mathematics education into the 21st century: Mathematics for living. Amman, Jordan. Ahlawat, K., and Da'jeh. H. (1996, December). Enhancing instructional quality through educational reform: Value- added after four years. Paper presented at the Forty-Third IECT World Assembly held in Amman, Jordan. Ahmed, A. 1987, Better Mathematics, HMSO, p. 24. Alessi, S. M., and Trollip, S. R. (2001). Multimedia for learning: Methods and development (3rd ed.). Needham Heights, MA: Allyn and Bacon. American Psychological Association. (1995). Learner-centered psychological principles: A framework for school redesign and reform (Report No. ED 411- 493). Washington, D.C.: American Psychological Association. Ames, C. (1992). Classrooms: Goals, structures, and student motivation. Journal of Educational Psychology, 84, 261–271. Ames, C. A. (1990). Motivation: What teachers need to know. Teachers College Record, 91(3), 409-422. Retrieved from: EBSCOhost on-line database (EBSCOhost, AN 9705111791). Apple, M. (1992). Do the standards go far enough? Power, policy, and practice in mathematics education. Journal for Research in Mathematics Education, 23, 5, 412-431. Aronson, Dennis T., and Briggs, J. (1983). Contributions of Gagné and Briggs to a prescriptive model of instruction, in Reigeluth, C.M. (1983) (e.d). Instructional design theories and models: An overview of their current Status. Hillsdale, NJ: Prentice-Hall. Artzt, A. F., and Armour-Thomas, E. (2003). Cognition and Instruction. Bahrick, H. P., Hall, L. K., and Dunlosky, J. (1993). Reconstructive processing of memory content for high versus low test scores and grades. Applied Cognitive Psychology, 7, 1-10. Baroody, A. J. (1998). Fostering children's mathematical power: An investigative approach to K8 mathematics instruction. Mahwah, NJ: Lawrence Erlbaum Associates.

202

Bennett, N., and Desforges, C. (1988). Matching classroom tasks to students’ attainments. Elementary School Journal, 88, 221–234. Berk, L., and Winsler, A. (1995). Scaffolding Children's Learning: Vygotsky and Early Childhood Education. Washington D.C.: National Association for the Education of Young Children. Biggs, J. B., and Moore, P. J. (1993). The process of learning. New York: Prentice Hall. Bonk, C. J., and Reynolds, T. H. (1997). Learner-centered Web instruction for higherorder thinking, teamwork, and apprenticeship. In B. Khan (Ed.), Webbased instruction(pp. 167-178). Englewood Cliffs, NJ: Educational Technology Publications. Borasi, R. (1996). Reconceiving mathematics instruction: A focus on errors. Norwood, NJ: Ablex Publishing Corporation. Borich, G. and Tombari, M (1997) Educational psychology: A contemporary approach. 2nd edition. NY: Addison Wesley. Bransford, J. D., Brown, A. L., and Cocking, R. R. (Eds.). (2000). How people learn: Brain, mind, experience, and school. Washington, DC: National Academy Press. Brown, A. L. (1987) Metacognition, executive control, self-regulation, and other even more mysterious mechanisms. In Weinert, F. E. and Kluwe, R. H. (eds.) Metacognition, motivation, and understanding. Hillsdale, N.J: Lawrence Erlbaum Associates. Brown, A. L., Campione, J. C., Reeve, R. A., Ferrara, R. A., and Palincsar, A. S. (1991). Interactive learning and individual understanding: The case of reading and mathematics. In Landsmann, L. T. (Ed.), Culture, schooling, and psychological development (pp. 136-170). Norwood, NJ: Ablex Publishing Co. Brown, P., and Goren, P. (1993). Ability, grouping and tracking: Current issues and concerns. Washington, DC: National Governors' Association. Brown, C., Hedberg, J., and Harper, B. (1994). Metacognition as a basis for learning support software. Performance Improvement Quarterly, 7, 2, 3-26. Brown. J. S. Collenis. A and Duguid. P. (1989). Situated cognition and the culture of learning. Educational Researcher 18. 32-42. Brown, A., L., and Palincsar, A. S., (1989). Guided, cooperative learning and individual knowledge acquisition. In L. B. Resnick (Ed.), Knowing, learning, and instruction: Essays in honor of Robert Glaser (pp. 393-451). Hillsdale, NJ: Lawrence Erlbaum Associates.

203

Bruner, J. (1960). The Process of Education. Cambridge, MA: Harvard University Press. Bruner, J. (1973). Going Beyond the Information Given. New York: Norton. Burron, B., James, M., and Ambrosio, A. (1993). The effects of cooperative learning in a physical science course for elementary/middle level preservice teachers. Journal of Research in Science Teaching, 30, 7, 697-707. Butler, D. L. (1995). Promoting strategic learning by post secondary students with learning disabilities. Journal of Learning Disabilities, 28, 170-190 Caplow, J., and Kardash, C. (1995). Collaborative learning activities in graduate courses. Innovative Higher Education, 19, 3, 207-221. Carpenter, T. A. and Lehere, R. (1999). Teaching and learning mathematics with understanding. In E. Fennema and T. A. Romberg (Eds), Classroom that promote mathematical understanding. Mahwah, NJ: Erlbaum. Carpenter, T. P., Franke, M. L., Jacobs, V. R., Fennema, E., and Empson, S. B. (1998). A longitudinal study of invention and understanding in children’s multidigit addition and subtraction. Journal for Research in Mathematics Education 29, 1, 3-20. Cecil, S. J., and Roazzi, A. (1994). The effects of context on cognition: Postcards from Brazil. In J. S. Sternberg and R. K. Wagner (Eds.), Mind in context: Interactionist perspectives on human intelligence (pp. 74-100). England: Academic Press. Chi, M., Bassok., M., Lewis, M., Reimann, P., and Glaser, R. (1989). Selfexplanations: How students study and use examples in learning to solve problems. Cognitive Science, 13, 145-182. Chi, M. T. H., and Glaser, R. (1985). Problem solving ability. In R. J. Sternberg (Ed.), Human abilities: An information processing approach (pp. 227-250). New York: W. H. Freeman and Company. Chi, M., de Leeuv, N., Chiu, M.-H., and Lavancher, C. (1994). Eliciting selfexplanations improves understanding. Cognitive Science, 18, 439 - 477. Chung, J. (2002). The Effect of the Availability of Technology on Teachers’ Use of Technology and Student Achievement on Standardized Tests. Unpublished Doctoral Dissertation, the Pennsylvania State University. Clark, R. C. (1995).21st Century Human Performance. Training, 6, 85-90. Coakes, J., Steed, G. (2001). SPSS. Analysis Without Anguish: Version 10.0 for Windows. Australia: John Wiley and Sons.

204

Cobb, P., Yackel, E., and Wood, T. (1995). The teaching experiment classroom. In P. Cobb and H. Bauersfeld (Eds.). The emergence of mathematical meaning; Interaction in classroom cultures (pp. 17-24). Hillsdale, NJ: Erlbaum. Costa, A. L. (1985). Developing Minds: A Resource Book of Teaching Thinking. Reston, Va: Association for Supervision and Curriculum Development. Cossey, R. (1997). Mathematical communication: Issues of access and equity. Unpublished doctoral dissertation, Stanford University. Cripps and McGilchrist (1999), in Mortimore (Ed) (1999: p47). Understanding Pedagogy and Its Impact on Learning. London: Paul Chapman. Cross, D.R. and Parts, S.G. (1988) Developmental and instructional analyses of children's metacognition and reading comprehension, Journal of Educational Psychology, 80, 131-142. Dansereau, D. F. (1988). Cooperative learning strategies. In C. E. Weinstein, E. T. Goetz, and P.A. Alexander (Eds.), Learning and study strategies: Issues in assessment, instruction, and evaluation (pp. 103-120). New York: Academic Press. Davidson, N. (1990).Cooperative Learning in Mathematics, New York: Academic Press Davis, E. A., and Linn, M. (2000). Scaffolding students' knowledge integration: Prompts for reflection in KIE. International Journal of Science Education, 22, 8, 819- 837. DeLay, R. (1996) Forming knowledge: Constructivist learning and experiential education. The Journal of Experiential Education, 19, 76-81. Dewey. J. (1961) Democracy and education. New York: Macmillan. Donn, S., and Taylor, A. (1992a). 3a focus group working paper, (TIMSS document ref. ICC313). Vancouver: International Coordination Centre. Donovan, M., Bransford, J. and Pellegrion, J. (1999). How people learn: Bridging research and practice. Washington, DC: National Academy Press. Doyle, W. (1988). Work in mathematics classes: The context of students' thinking during instruction. Educational Psychologist, 23, 167–180. Driscoll, M.P. (1998). Web-based training: Using technology to design adult learning experiences. San Francisco, CA: Jossey-Bass/Pfeiffer. Driscoll, M.P. (1994). Psychology of learning for instruction. Boston: Allyn and Bacon.

205

Duda, J. L., and Nicholls, J. G. (1992). Dimensions of achievement motivation in schoolwork and sport. Journal of Educational Psychology, 84, 290–299. Dweck, C. (1986). Motivational processes affecting learning. American psychologist, 41, 1040-1048. Elawar, C. M. (1992). Effects of teaching metacognitive skills to students with low mathematics ability. Teaching and Teacher Education, 8, 109 – 21. Ertmer, P. A., and Newby, T. J. (1996). The expert learner: Strategic, self-regulated, and reflective. Instructional Science, 24, 1-24. Feltovich, P. J., Spiro, R. J., Coulson, R. L., and Feltovich, J. (1996). Collaboration within and among minds: Mastering complexity, individuality and in groups. In T. Koschmann (Ed.), CSCL: Theory and practice of an emerging paradigm. Mahwah, NJ: Lawrence Erlbaum Associates. Flavell, J. H. (1963). The developmental psychology of Jean Piaget. New York: D. Van Nostrand. Flavell, J. H. (1976). Metacognitive aspects of problem solving. In L. B. Resnick (Ed.), The nature of intelligence. Hillsdale, NJ: Erlbaum. Flavell, J. H. (1977). Cognitive development. Englewood Cliffs, NJ: Prentice-Hall. Flavell, J. H. (1979). Metacognition and cognitive monitoring: A new area of cognitive-developmental inquiry. American Psychologist, 34, 906-911. Flavell, J. H., Friedrichs, A.G. and Hoyt, J.D. (1970) Developmental changes in memorisation processes. Cognitive Psychology, 1, 323-340 Fosnot, C. (1996). Constructivism: A Psychological theory of learning. In C. Fosnot (Ed.) Constructivism: Theory, perspectives, and practice, (pp.8-33). NY: Teachers College Press. Frankenstein, M. (1990). Incorporating race, gender, and class issues into a critical mathematical literacy curriculum. Journal of Negro Education, 59, 3, 336347. Gagne, R. (1985). The Conditions of Learning (4th ed.). New York: Holt, Rinehart and Winston . Garner, R., and Alexander, P. A. (1989). Metacognition: Answered and unanswered questions. Educational Psychologist, 24, 143-158. Garofalo, J., and Lester, F. K. (1985). Metacognition, Cognitive Monitoring, and Mathematical Performance. Journal for Research in Mathematics Education 16, 3, 163-176

206

Garrison, D. R. (1997). Self-directed learning: Toward a comprehensive model. Adult Education Quarterly, 48(1), 18-34. Retrieved from: EBSCOhost on-line database (EBSCOhost, AN 1486). Gavelek, J. R., and Raphael, T. E. (1985). Metacognition, instruction, and the role of questioning activities. In D. L. Forrest-Pressley, G. E. MacKinnon, and T. G. Waller. (Eds.), Metacognition, cognition, and human performance (pp. 103-136). Orlando, FL: Academic Press. Gick, M. L. (1986). Problem solving strategies. Educational Psychologist, 21, 1 and 2, 99-120. Golinkoff, R. A. (1976). A comparison of reading comprehension processes in good and poor comprehenders. Reading research quarterly, 11, 623-659. Gordon, J. (1996). Tracks for learning: Metacognition and learning technologies. Australian Journal of Educational Technology, 12, 1, 46-55. Retrieved from the World Wide Web: http://cleo.murdoch.edu.au/gen/aset/ajet/ajet12/wi96p46.html Graesser, A. C., and Person, N. K. (1994). Question asking during tutoring. American Educational Research Journal, 31, 1, 104-137. Greene, B. A., and Land, S. M. (2000). A qualitative analysis of scaffolding use in a resource-based learning environment involving with the World Wide Web. Journal of Educational Computing Research, 23, 2, 151-180. Hacker, D. J. (1998). Metacognition: Definitions and Empirical Foundations. In D. J. Hacker, J. Dunlosky, and A. C. Graesser (Eds.), Metacognition in Educational Theory and Practice (pp. 1-23). Lawrence Erlbaum Associates. Hagarty, M., Mayer, R. E., and Monk, C. A. (1995). Comprehension of arithmetic word problems: A comparison of successful and unsuccessful problem solvers. Journal of educational psychology, 87, 18-32. Hair, F. J., Anderson, E., Tatham, L., and Black, C. (1998). Multivariate Data Analysis (5 th edition). New Jersey: Prentice Hall. Halpren, D. (1996). Thought and knowledge: An introduction to critical thinking. Mahwah, NJ: Lawrence Erlbaum Associates, 58-59. Hanley, S. (1994). On constructivism. Available at: http:// www. Inform. Umd.edu/UMS+ State/UMD Projects/MCTP/Essays/Constructivism.txt. Hein, G. (1991, October). Constructivist learning theory. International committee of museum educators, Conference Jerusalem.

207

Hiebert, J., and Carpenter, T. P. (1992). Learning and teaching with understanding. In D. A. Grouws (Ed.), Handbook of research on mathematics teaching and learning (pp. 65–97). New York: Macmillan. Hoek, D., van den Eden, P., and Terwel, J. (1999). The effects of integrated social and cognitive strategy instruction on the mathematics achievement in secondary education. Learning and Instruction, 9, 427-448. Hong, N. S. (1998). The relationship between well-structured and ill-structured problem solving in multimedia simulation. Unpublished doctoral dissertation, The Pennsylvania State University, University Park, PA. Inhelder, B., and Piaget, J. (1958). The growth of logical thinking from childhood to adolescence. New York: Basic Books. Inhelder, B., and Piaget, J. (1964). The Early Growth of Logic in the Child. NY: Harper and Row. Innabi, Hanan; Kaisee, and Hind (1995). Learning achievement of basic cycle students in mathematics. Amman: NCHRD. Ismail, H-N,. (1999). Learning within scripted and non-scripted peer tutoring sessions: The Malaysian context. Unpublished Doctoral Dissertation, Indiana University. Jacobs, J. E., and Paris, S. G. (1987). Children's metacognition about reading: Issues in definition, measurement, and instruction. Educational Psychologist, 22, 255-278. Jacobson, R. (1998). Teachers improving learning using metacognition with selfmonitoring learning strategies. Education, 118, 4, 579-589. Jbeili, I. M. (1999). The effects of instructional computer on direct and delayed achievement of fifth grade students in mathematics. Unpublished Master Thesis, Yarmouk University. Johnson, D. W. (1981). Student-student interaction: the neglected variable in education. Educational Researcher, 2, 5-10. Johnson, D. W., Johnson, R. T., and Smith, K. A. (1991). Active learning: Cooperation in the college classroom. Edina, MN: Interaction Book Company. Johnson, D., and Johnson, R. (1999). Learning together: Cooperative, competitive, and individualistic learning. Massachusetts: Allyn and Bacon. Johnson, D.W. and R.T. Johnson. (1987). Learning Together and Alone (2nd Ed). Englewood Cliffs, NJ: Prentice Hall.

208

Johnson, R. T., Johnson, D. W., and Stanne, M. B. (1986). Comparison of computer assisted cooperative, competitive, and individualistic learning. American Educational Research Journal, 23, 3, 382-392. Jonassen, D. H. (1991) Objectivism Versus Constructivism: Do We Need a New Philosophical Paradigm? Educational Technology Research and Development, 39, 3, 5 – 14. Jonassen, D. H. (1997). Instructional design models for well-structured and illstructured problem-solving learning outcomes. Educational Technology: Research and Development, 45, 1, 65-94. Kamii, C. (2000). Young Children Reinvent Arithmetic. 2nd ed. NY: Teachers College Press. Kaur, B. and Pereira, M. L. (2000). TIMMS – Performance of Singapore secondary students, Part B: Proportionality, Measurement, Fractions, and Number Sense. Journal of Science and Mathematics Education in Southeast Asia, 23, 1, 54-70. Kilpatrick, J., Swafford, J., and Findell, B. (2001). Adding It Up: Helping Children Learn Mathematics. Washington, DC: National Academy Press. King, A. (1991a). Effects of training in strategic questioning on children's problemsolving performance. Journal of Educational Psychology, 83, 3, 307-317. King, A. (1991b). Improving lecture comprehension: Effects of a metacognitive strategy. Applied Cognitive Psychology, 5, 331-346. King, A. (1992). Facilitating elaborative learning through guided student-generated questioning. Educational Psychologist, 27, 1, 111-126. King, A., and Rosenshine, B. (1993). Effect of guided cooperative questioning on children's knowledge construction. Journal of Experimental Education, 61(2), 127-148 King, A. (1989). Effects of self-questioning training on college students’ comprehension of lectures. Contemporary Educational Psychology, 14, 366 – 381. Kitchner, K. S. (1983). Cognition, metacognition, and epistemistic cognition: A threelevel model of cognitive processing. Human Development, 26, 222-232. Kohn, A. (1992). Resistance to cooperative learning: Making sense of its deletion and dilution. Journal of Education, 174, 2, 38-56.

209

Kramarski, B. (2000). The effects of different instructional methods on the ability to communicate mathematically. In Tadao Nakahara and Masataka Koyama (eds.). Education, proceedings of the 24th Conference of the International Group for the Psychology of Mathematics Mathematical Reasoning. Hiroshima University, Hiroshima, Japan, 1, 167 – 171. Kramarski, B., Mevarech, Z. R., and Lieberman, A. (2001). Effects of multilevel versus unilevel metacognitive training on mathematical reasoning. Journal of educational research, 94, 292-301. Kramarski, B., Mevarech, Z. R., and Arami, M. (2002). The effects of metacognitive instruction on solving mathematical authentic tasks. Educational Studies in Mathematics, 49, 225 – 250. Landa, L. (1976). Instructional regulation and control: cybernetics, Algorithmization, and heuristics in education. Englewood cliffs, NJ: Educational Technology Publications. Landa, L. (1983). The algo-heuristic theory of instruction. In Reigeluth, C. M. (1983). Instructional- design theories and models. An overview of their current status. Hillsdale NJ: PrenticeHall. Lawson, A. E. (1995). Science teaching and the development of thinking. California: Wadsworth Publishing Company. Leder, G. C. (1987). Teacher student interaction: A case study. Educational Studies in Mathematics, 18, 255-271. Lepper, M. R. (1988). Motivational considerations in the study of instruction. Cognition and Instruction, 5, 289–309. Lesh, R. (1985). Processes, skills, and abilities needed to use mathematics in everyday situations. Education and Urban Society, 17, 4, 439-446. Lin, X., Hmelo, C., Kinzer, C. K., and Secules, T. J. (1999). Designing technology to support reflection. Educational Technology: Research and Development, 47, 3, 43-62. Lin, X., and Lehman, J. D. (1999). Supporting learning of variable control in a computer-based biology environment: Effects of prompting college students to reflect on their own thinking. Journal of Research in Science Teaching, 3, 7, 837-858. Manion, V. (1995). The benefits of peer collaboration on strategy use, metacognitive causal attribution, and recall. Unpublished Doctoral Dissertation, Indiana University.

210

Masui, C., and De Corte, E. (1999). Enhancing learning and problem solving skills: Orienting and self-judging, two powerful and training learning tools. Learning and Instruction, 9, 6, 517-542. Mathematics framework for California public schools (1999). Kindergarten through grade twelve. Sacramento: California Department of Education Mayer, R. (1996). Students as information processors: Legacies and limitations of educational psychology's second metaphor. Educational Psychologist, 31, 151 - 161. Meichenbaum, D. (1976). Cognitive factors as determinants of learning disabilities; A cognitive function approach. In Knights, R., and Bakker, D. (Eds), The neuropsychology of learning disorders: Theoretical approaches. Baltimore: University Park Press. Mevarech, Z.R., and Light, P.H. (1992). Peer-based interaction at the computer: looking backward, looking forward. Learning and Instruction, 2, 275-280. Mevarech, Z. R., and Kramarski, B. (1997). IMPROVE: A multidimensional method for teaching mathematics in heterogeneous classrooms. American Educational Research Journal, 34, 2, 365-395. Mevarech, Z. R. (1999). Effects of metacognitive training embedded in cooperative settings on mathematical problem solving. The Journal of Educational Research, 92, 4, 195 – 205. Mevarech, Z. R., and Kramarski, B. (in press). The effects of metacognitive training versus worked-out examples on students' mathematical reasoning. Journal of Educational Psychology. Middleton, J. A. (1993a). An analysis of the congruence of teachers' and students' personal constructs regarding intrinsic motivation in the mathematics classroom. Doctoral dissertation, University of Wisconsin-Madison. Dissertation Abstracts International, 53, 31-50A. Middleton, J. A., Littlefield, J., and Lehrer, R. (1992). Gifted students' conceptions of academic fun: An examination of a critical construct for gifted education. Gifted Child Quarterly, 36, 38-44. Montague, M., and Bos, C. S. (1990). Cognitive and metacognitive characteristics of eighth-grade students' mathematical problem solving. Learning and Individual Differences, 2, 371-388. Moore, M. G., and Kearsley, G. (1996). Distance education: A systems view. Belmont, CA: Wadsworth Publishing Company. Mugny, G., and Doise, W. (1978). Socio-cognitive conflicts and structure of individual and collective performances. European Journal of Social Psychology, 8, 1181-1192.

211

National assessment of educational progress (2000). National assessment governing board. Mathematics framework for the 1996 and 2000 National assessment of educational progress. Washington, DC: Author. Available: http://www.nagb.org. National Council of Teachers of Mathematics. (1989). Curriculum and evaluation standards for teaching mathematics. Reston, VA: Author. National Council of Teachers of Mathematics. (1991). Professional standards for teaching mathematics. Reston, VA: Author. National Council of Teachers of Mathematics. (1995). Assessment standards for school mathematics. Reston, VA: Author. National Council of Teachers of Mathematics. (2000). Principles and standards for school mathematics. Reston, VA: Author. National Research Council. (1989). Everyday counts: A report to the nation on the future of mathematics education. Washington, DC: National Academy Press. Available: http://books.nap.edu/catalog/5777.html. National Research Council. (1998). High school mathematics at work: Essays and Examples for the education of all students. Washington, DC: National Academy Press. Available: http://books.nap.edu/catalog/5777.html. North Central Regional Educational Laboratory. (1995). Metacognitive questions. Available: http://www.ncrel.org O’deh, A., Malkawi, F. (1992). The scientific research basics in education and human sciences, research components and their statistical data analysis. Jordan, Irbid: Yarmouk University. OECD: 2000, Measuring Students Knowledge and skills, The PISA 2000 assessment of reading, mathematical and scientific literacy. Osman, M. E and Hannafin, M. J. (1994). Effects of advance questioning and prior knowledge on science learning. Journal of Educational Research, 88, 1, 5-13. Ossont, D. (1993). How I use cooperative learning. Science Scope, 16, 8, 28-31. Palincsar, A. S. (1986). The role of dialogue in providing scaffolded instruction. Educational Psychologist, 21, 1 and 2, 73-98. Palincsar, A. S., and Brown, A. L. (1984). Reciprocal teaching of comprehensionfostering and comprehension-monitoring activities. Cognition and Instruction, 2, 117-175

212

Palincsar, A. S., Brown, A. L., and Martin, S. M. (1987). Peer interaction in reading comprehension instruction. Educational Psychologist, 22, 3 and 4, 231-253. Palincsar, A. S., and Brown, A. L. (1989). Instructions for self-regulated reading. In L. B. Resnick and L. E. Klopfer (Eds.), Toward the thinking curriculum: Current cognitive research (pp. 19-39). Washington, DC: Association for Supervision and Curriculum Development. Paris, S. G. and Winograd, P. (1990). How metacognition can promote academic learning and instruction. In B. F. Jones and L. Idol (Eds.), Dimensions of thinking and cognitive instruction (pp. 15-51), Hillsdale, NJ: Lawrence Erlbaum. Pelligrini, A. D., Galda, L., and Flor, D. (1996). Relationships, Individual Differences and Children's Use of Literate Language. Unpublished manuscript. Perkins, D. N., and Grotzer, T. A. (1997). Teaching intelligence. American Psychologist, 52, 10, 1125-1133. Peterson, P. L., Swing, S. R., Stark, K. D., and Waas, G. A. (1984). Students' cognitions and Time on task during mathematics instruction. American Educational Research Journal, 21: 487-515. Peterson, P.L., and Swing, S. R., Braverman, M. T., and Buss, R. (1982). Students' aptitudes and their reports of cognitive processes during direct instruction. Journal of Educational Psychology, 74, 4, 535-47. Petraglia. J. (1998). The real world on a short leash: The (Mis) Application of constructivism to the design of educational technology. ETR and D, 46, 53-65. Piaget, J. (1970). The Science of Education and the Psychology of the Child. New York: Grossman. Piaget, Jean. (1971). Biology and Knowledge. Chicago: University of Chicago Press. Piaget, J., and Szeminska, A. (1965). The Child's Conception of Number. New York: W. W. Norton and Co. Pollack, H. (1997). Solving problems in the real world. In L. A. Steen (Ed.), Why numbers count: Quantitative literacy for tomorrow's America (pp. 91 - 105). New York: College Entrance Examination Board. Pressley, M., and McCormick, C. B. (1987). Advanced educational psychology for educators, researchers, and policy makers. New York: HarperCollins. Qin, Z., Johnson, D. W., and Johnson, R. T. (1995). Cooperative versus competitive efforts and problem solving., Review of Educational Research 65, 129 – 144. Resnick, L. B. (1987). Education and learning to think. Washington, DC: National Academy Press.

213

Rogoff, B. (1990). Apprenticeship in thinking: Cognitive development in social context. NY: Oxford University Press. Romberg, T. (1988). Comments: NCTM’s curriculum and evaluation standards. Teacher College Record, 100, 1, 8-21. Rosenshine, B. and Meister, C. (1993). The Use of Scaffolds for Teaching HigherLevel Cognitive Strategies. In Woolfolk, A.K. (Ed.). Readings and Cases in Educational Psychology, 5th ed., Boston: Allyn and Bacon. Russell, S. J. (2000). Developing computational fluency with whole numbers in the elementary grades. In Ferrucci, B. J. and Heid, M. K. (eds). Millenium focus issue: Perspectives on principles and standards. The New England Math Journal. XXXII, 2, 40-54 Keene, NH: Association of Teachers of Mathematics. Ryan, E. B. (1981). Identifying and remediating failures in reading comprehension: Toward an instructional approach for poor comprehension. In T. G. Waller and G. E. MacKinnon (Eds.), Advances in reading research. New York: Academic Press. Salomon, G., Globerson, T., and Guterman, E. (1989). The computer as a zone of proximal development: Internalizing reading-related metacognitions from a reading partner. Journal of Educational Psychology, 81, 4, 620-627 Sapon-Shevin, M., and Schniedewand, N. (1992). If cooperative learning's the answer, what are the questions? Journal of Education, 174, 2, 10-37. Scandura, J.M. (1977). Problem Solving: A Structural/Process Approach with Instructional Applications. NY: Academic Press Scandura, M. (1983). Instructional strategies based on the structural learning theory. In Reigeluth, C. M. (1983). Instructional- design theories and models. An overview of their current status. Hillsdale NJ: Prentice-Hall Scardamalia, M., Bereiter, C., and Steinbach, R. (1984). Teach ability of reflective processes in written composition. Cognitive Science, 8, 173-190. Schoenfeld, A. H. (1985) Mathematical Problem-Solving. Lawrence Erlbaum Associates. Schoenfeld, A. H. (1987). What's all the fuss about metacognition? In A. H. Schoenfeld (Ed.), Cognitive science and mathematics education (pp. 189-215). Hillsdale, NJ: Lawrence Erlbaum Associates. Schoenfeld, A. H. (1992). Learning to think mathematically: Problem solving, metacognition, and sense making in mathematics. In D. A. Grouws (Ed.), Handbook of research on mathematics teaching and learning (pp. 334-371). New York: Macmillan.

214

Schraw, G. (1998). On the development of adult metacognition. In M. C. Smith and T. Pourchot (Eds.), Adult learning and development: Perspectives from education psychology (pp. 89-106). Mahwah, NJ: Lawrence Erlbaum Associates. Schraw, G., and Dennison, R. S. (1994). Assessing Metacognitive Awareness. Contemporary Educational Psychology, 19, 460-475. Shachar, H. and Sharan, S. (1994) Talking, relating and achieving: Effects of cooperative learning and whole-class instruction, Cognition and Instruction, 12, 4, 313-353. Shulman, S. (1973). Psychological Controversies in the Teaching of Science and Mathematics. In Crosswhitein, F. J., Higgins, J. L., Osborne, A. R., and Shumway, R. J. (Eds.). Teaching Mathematics: Psychological Foundations (pp. 3-18). Worthington, Ohio: Charless A. Jones Publishing Company. Silver, E. A. (1987). "Foundations of Cognitive Theory and Research for Mathematics Problem Solving Instruction." In A. H. Schoenfeld, ed., Cognitive Science and Mathematics Education. Hillsdale, NJ: Lawrence Erlbaum Associates Simpson, M. L. (1984). The status of study strategy instruction: implications for classroom teachers. Journal of Reading, November, 136 – 42. Sinnott, J. D. (1989). A model for solution of ill-structured problems: Implications for everyday and abstract problem solving. In J. D. Sinott (Ed.), Everyday problem solving: Theory and application (pp. 72-99). New York: Praeger. Skinner, B. (1968). The technology of teaching. New York: Appleton. Slavin, R. E. (1996). Research on cooperative learning and achievement: What we know, what we need to know. Contemporary Educational Psychology, 21, 4369. Slavin, R.E. (1994). Educational Psychology: Theory and Practice. Needham Heights: Allyn and Bacon. Slavin, R.S. (1987). Cooperative learning and the cooperative school. Educational Leadership, 45, 4, 7-13. Smith, P. L., and Ragan, T. J. (1993). Instructional design. Upper Saddle River, New Jersey: Prentice-Hall. Stacey and Kay. (1992). Mthematical problem solving in groups – are two heads better than one? Journal of Mathematical Behavior 11, 3, 261 – 275. Stein, M. K., Grover, B. W. and Henningsen, M. (1996). Building student capacity for mathematical thinking and reasoning: An analysis of mathematical tasks used in reform classroom. American Educational Research Journal, 33, 2, 455-488. Sternberg, R. J. (1986b). Intelligence applied. New York: Harcourt Brace Jovanovich, Publishers.

215

Sternberg, R. J., and Rifkin, B. (1979). The development of analogical reasoning processes. Journal of Experimental Child Psychology, 27, 195-232. Stevens, J. (1986). Multivariate statistics for the social science. Hillsdale, New Jersey: Lawrence Erlbaum Associates. Thorndike, E. (1932). The Fundamentals of Learning. New York: Teachers College Press. Thornton, C. A., and Toohey, M. A. (1985). Basic math facts: Guidelines for teaching and learning. Learning Disabilities Focus, 5, 1, 44-57. Tiene, D., and Ingram, A. (2001). Exploring Current Issue in Educational Technology. N Y: Ridge and Madison. TIMSS-R International study center (1999). International Mathematics Report. Findings from IEA's repeat of the Third international mathematics and science study at the eighth grade. Chestunt Hill, MA: Boston College. Tuckman, B. W. (1999). Conducting educational research. 5th edition. Orlando: Harcourt Brace College Publishers. Van Zee, E., and Minstrell, J. (1997). Using questioning to guide student thinking. The Journal of the Learning Sciences, 6, 2, 227-269. von Glasersfeld, E. (1995). Radical constructivism: A way of knowing and learning. Washington, DC: The Falmer Press. Voss, J. F., and Post, T. A. (1988). On the solving of ill-structured problems. In M. H. Chi, R. Glaser and M. J. Farr (Eds.), The nature of expertise (pp. 261-285). Hillsdale, NJ: Lawrence Erlbaum Associates. Voss, J. F., Wolfe, C. R., Lawrence, J. A., and Engle, R. A. (1991). From representation to decision: An analysis of problem solving in international relations. In R. J. Sternberg and P. A. Frensh (Eds.), Complex problem solving. Hillsdale, NJ: Lawrence Erlbaum Associates. Vygotsky. L.S. (1978). Mind in society: the development of higher psychological processes. Cambridge, MA: Harvard University Press. Webb, N. M. (1989a). Critical distinctions among three approaches to peer education. International Journal of Educational Research, 13, 9-19. Webb, N. (1989b). Peer interaction and learning in small groups. International Journal of Educational Research, 13, 21-40. Webb, N. M. (1991). Task – related verbal interaction and mathematics learning in small groups. Journal for Research in Mathematics Education, 22, 366 – 389. Webb, N. M. (1982). Group composition, group interaction and achievement in cooperative small groups. Journal of Educational Psychology, 74, 475-484.

216

Webb, N. M., and Palincsar, A. S. (1996). Group processes in the classroom. In D. C. Berliner and R. C. Calfee (Eds.), Handbook of educational psychology (pp. 841-873). New York: Simon and Schuster Macmillan. Weinstein, C. E., Meyer, D. K. and Stone, G. V. M. (1994). Teaching Students How to Learn. In W. J. McKeachie, N. Chism, R. Menges, M. Svinicki and C. E. Weinstein (Eds.), Teaching tips (9th Ed.) (pp. 359-367). Lexington, Toronto: D.C. Heath and Company. Wertheimer, M. (1959). Productive Thinking (Enlarged Ed.). New York: Harper and Row. White, L. F. (1998). Motivating students to become more responsible for learning. College Student Journal, 32, 2, 190-197. Retrieved from EBSCOhost on-line database (EBSCOhost, AN 759257). Wiig, E., and Semel, E. (1984). Language assessment and intervention for the learning disabled child. Columbus, OH: Merrill. Wilkins and Jesse, L. M. (2000). Preparing for the 21st century: The status of quantitative literacy in the United States. School Science and Mathematics, 100, 8, 405-418. Wineburg, S. S. (1998). Reading Abraham Lincoln: An expert-expert study in the interpretation of historical texts. Cognitive Science, 22, 319-346. Wittrock, M. C. (1986). Students' thought processes. In M. C. Wittrock (Ed.), Handbook of research on teaching (3rd ed.). New York: Macmillan. Wong, B. Y. (1985). Self-questioning instructional research: A review. Review of Educational Research, 55, 227-268. Wood, D. J., Bruner, J. S., and Ross, G. (1976). The role of tutoring in problem solving. Journal of Child Psychology and Psychiatry, 17, 89-100. Xun, G. (2001). Scaffolding students problem-solving processes on an ill-structured task using question prompts and peer interactions. Unpublished Doctoral Dissertation, the Pennsylvania State University.

217