You are on page 1of 19

JOURNAL OF RESEARCH IN SCIENCE TEACHING VOL. 27, NO. 4, PP.

315-333 (1990)

BUILDING AN ORGANIZED KNOWLEDGE BASE:


CONCEPT MAPPING AND ACHIEVEMENT
IN SECONDARY SCHOOL
PHYSICS*

WILLIAM J. PANKRATIUS
College of Education, University of Nevada,
Las Vegas, Nevada 89154

Abstract
Direct teaching of problem-solving methods to high school physics students met with little
success. Expert problem solving depended upon an organized knowledge base. Concept mapping
was found to be a key to organizing an effective knowledge base. The investigation of the effect
of the degree of concept mapping on achievement was the purpose of this study. Six intact high
school physics classes, taught by this investigator, took part in the study. Two classes were
control groups and received standard instruction. Four classes received six weeks of concept-
mapping instruction prior to the unit under study. Two of these four classes were the low-level
treatment group and were required to submit concept maps at the conclusion of the instruction.
The other two classes were the high-level treatment group and were required to submit concept
maps at the beginning and at the conclusion of the unit under study. One class from each
treatment group took a pretest prior to instruction. An analysis of the posttest results revealed
no pretest sensitization. A one-way analysis of covariance indicated a significant main effect
for the treatment level at the p < 0.05 level. A pair of single-dfcomparisons of the adjusted
treatment means resulted in significant differences (p < 0.05) between the control group and
the average of the treatment means as well as between the two experimental groups. It can be
concluded that for this sample (upper-middle-classhigh school physics students) mapping concepts
prior to, during, and subsequent to instruction led to greater achievement as measured by posttest
scores.

Introduction
A principal goal of physics teaching has been the improvement of the ability of
students to solve problems. Instruction in problem-solving methods has produced mixed
results. True (1974), Richardson (1981), Lin (1982), and Finkel (1983) have fourtd
that the teaching of problem solving methods failed to yield statistically significant

* This manuscript is based on a doctoral dissertation by the author, Georgia State University, 1987.

0 1990 by the National Association for Research in Science Teaching


Published by John Wiley & Sons, Inc. CCC 00224308/90/040315-19$04.00
316 PANKRATIUS

problem solving success. Lin (1982) suggested that students tended to use that method
which worked in the past. Problem solving, according to Greeno (1980), was based
on knowledge. Greeno’s advice on teaching students to solve a class of problem was:
“first analyze the knowledge that they need in order to solve that class of problems,
and then carry out instruction that will result in their acquisition of the required
knowledge” (Greeno, 1980, p. 13).
In order to determine what constituted an adequate knowledge base for the physics
problem solver, the role of the expert problem solver was examined. Larkin and Reif
(1979, p. 200) found that experts organized knowledge into “coherent methods, rather
than merely into individual principles or equations.” Chi, Feltovich, and Glaser (1981)
discovered that experts “categorize physics problems by the underlying physics princi-
ples, a kind of ‘deep structure,’ whereas novices categorize problems by the ‘surface
structure’ ” (p. 134). Chi et al. related the categorization and “its associated knowledge
.,
within the knowledge base” as constituting a “schema” (p. 135).
Novak ( 1976) began his “schema of conceptual hierarchy” (concept map) with
the most general, most inclusive concept at the top, and proceeded downward through
the “progressive differentiation of higher order concepts” to less general concepts and
specific examples (p. 509). Novak’s concept map functioned as a link between what
the learner already knew and the learning that was to come. The concept map, drawn
by the learner, became a snapshot of that learner’s mind as new concepts were worked
into his or her preexisting conceptual scheme.
By 198 1 Novak and his colleagues had refined the concept map to include linking
words that indicated relationships between concepts. Novak termed this relationship
between two concepts a proposition. These propositions, along with the levels of
hierarchy, reflected the extent of differentiation of the concepts. Valid cross links
between concepts found on different branches of the concept map were seen by Novak
(1984) to be “good indicators of integrative reconciliation of meanings” (p. 61 1).
A concept map, according to Novak (1984, p. 609), must have (a) a hierarchical
structure, (b) relationships between concepts identified, and (c) more specific concepts
subsumed under more inclusive concepts. A concept map became a teaching, or
learning, device that demanded that a student become actively involved in forming a
coherent, structured, integrated, easily accessible knowledge base anchored on a central
concept.
Concepts were not learned in isolation-they were placed in the students’ conceptual
network. The mapping of concepts served to externalize concepts and improved the
students’ conceptual scheme. Students’ concept maps changed dynamically as their
view of the universe was modified and refined.
Concept maps have been used to develop curriculum (Clibum, 1986), as instructional
devices (Stewart, Van Kirk & Rowell, 1979), for lecture preparation and laboratory
reports (Ault, 1985), and as evaluation tools (Novak, 1981). A concept map became
8 two-dimensional hierarchical organization of the structure of a discipline, a unit of
study, or even a paragraph. The various uses of concept maps are illustrated in Appendix
A: A concept map of concept maps.
Novak (1976) regarded problem solving as a process of meaningful learning-
building a knowledge base. The problem solver “meaningfully internalizes new elements,
thus differentiating some concepts further and forming new associations among subordinate
and/or superordinate concepts” (p. 500). Effective problem solving, which required
an efficient knowledge base, then contributed toward reinforcing that knowledge base.
BUILDING AN ORGANIZED KNOWLEDGE BASE 317

Several descriptive studies (Eylon & Reif, 1979; Stewart, Van Kirk & Rowell,
1979; Moreira, 1979; Champagne, Klopfer, Desena & Squires, 1981) detailed the
usefulness of internal knowledge organization and concept mapping, but there was a
shortage of studies that measured the effectiveness of concept mapping on achievement.
Bousquet (1982) conducted a study of three versions of concept mapping among
117 students in a college natural-resources management course. Concept-mapping skill
predicted subsequent achievement on a posttest; however, there was no significant
difference among the three treatment groups on the posttest scores.
Novak, Gowin, and Johansen (1983) studied the effect of concept mapping and
the use of Gowin’s Vee (a heuristic device that represented the structure of knowledge
involved in an activity) on seventh- and eighth-grade science students. The “experimental
classes demonstrated superiority in problem solving performance on novel problems
after less than six months of instruction” (p. 638). Their study suggested that concept
mapping required different abilities than those measured on standardized tests.
Eylon and Reif (1984) in a further study on internal knowledge organization
investigated students “who had acquired knowledge of a particular physics argument
organized in hierarchical form” (p. 38). These students “performed various recall and
problem solving tasks better than [students] who had acquired the same knowledge
organized at a single level” (p. 38).
Fraser and Edwards (1985) conducted a one-shot intervention treatment to deter-
mine the effect of concept mapping on student achievement in traditional classroom
tests. They found evidence that suggested that the students who scored highest on end-
of-unit tests had “high level[s] of concept mapping mastery” (p. 164). Lehman, Carter,
and Kahle (1985) conducted a rigorous study involving ten secondary school biology
classes. The study compared concept-mapping and [Gowin’s] Vee mapping heuristics
to outlining. Their results “suggest[ed] a tendency toward an effect of the experimental
technique” (p. 670).
Although concept mapping had ample theoretical support from the fields of cognitive
science and science education, there remained a large gap between concept-mapping
instruction and reported statistically significant achievement gains. It appeared that a
rigorous study, based on the efforts of Lehman, Carter, and Kahle (1985), and restricted
to different levels of concept mapping, would prove helpful in assessing achievement
changes that resulted from concept mapping.

Statement of the Problem


Successful problem solving required an extensive and well-organized knowledge
base. Concept mapping seemed to be an effective and efficient means to organize a
knowledge base. Problem solving, an organized knowledge base, and proper conceptual
links were thus assumed to be valid components of achievement. This study then
focused on the relationship between student concept mapping and achievement.
The main question under study then became

(Q-1) Are there any differences in achievement between groups that receive
instruction in concept mapping and groups that receive standard instruction?

A related question arose when concept mapping was considered as a form of


pretest and as a continuing instructional device throughout the unit:
318 PANKRATIUS

(4-2) Are there any differences in achievement between groups that map
concepts prior to, during, and subsequent to instruction and groups that concept
map subsequent to instruction only?

These questions produced the following null hypotheses:

H(0)l:There is no statistically significant difference in achievement between


groups that receive instruction in concept mapping and groups that receive standard
instruction.
H(0)2: There is no statistically significant difference in achievement between
groups that map concepts prior to, during, and subsequent to instruction and groups
that map concepts subsequent to instruction only.

The following research, or alternate, hypotheses were then developed:

H(R) 1 : Groups that receive concept-mapping instruction score significantly


higher on an achievement test than groups that receive standard instruction.
H(R)2: Groups that map concepts prior to, during, and subsequent to instruction
score Significantly higher on an achievement test than groups that map concepts
subsequent to instruction only.

Methodology

Subjects

A beginning pool of 145 high school physics students in six class periods was
reduced to a sample of 87 as a result of missing data. There was some overlap of
students for whom data was missing: (a) 39 failed to take the posttest as scheduled;
(b) 21 did not take the Scholastic Achievement Test, which was used as the covariant;
(c) 11 did not submit a concept map; (d) four did not take the pretest; and (e) one
student dropped the course.
The sample was primarily composed of male seniors (66.7%).Female seniors
made up 20.9% of the sample. Male juniors accounted for 3.4%.The average age of
the subjects was 17.
A school and community survey conducted in April, 1987 revealed that

1. Seventy-four percent of the students were born outside of the state of


Georgia.
2. Ninety-three percent of the students planned to attend college or some other
post secondary education.
3. The majority of the students were from upper-middle-class homes.
4. Sixty-one percent of their parents considered themselves professional.
5 . Thirty-eight percent of their parents considered themselves management.

lnstru~ent
The instrument chosen for measuring achievement for the unit under study consisted
of 30 items selected from the Ontario Assessment Instrument Pool: Physics-Senior
Division (OAIPPSD, 1981). The OAIPPSD was the largest collection of assessment
BUILDING AN ORGANIZED KNOWLEDGE BASE 319

items available in the field of secondary school physics. The items were designed to
meet the “aims, goals, and objectives of Senior Division Physics” (p. 1). The items
were collected or developed by the Science Teachers’ Association of Ontario and the
Department of Measurement and Evaluation of the Ontario Institute for Studies in
Education. Some of the items formed part of the Ontario Physics Achievement Tests.
The items were designed to be “used by teachers to evaluate students’ achievement
of a specific objective” (p. 2). Each item was classified as to content and behavior.
The items selected for this study were matched by behavioral objectives to the objectives
of the unit under study. Content validity of the instrument was assumed based on the
selection process of the test items. The absence of a standardized published instrument
(for the unit under study) resulted in no criterion-related validity for the assessment
instrument. Items from OAIPPSD have been used in this author’s physics courses for
pretests, posttests, and final examinations. There has been a degree of reliability from
test to test (Pearson r > 0.60). The items chosen for this assessment instrument may
be obtained from the author.
A split-half reliability analysis was performed on the posttest results subsequent
to an item-by-item rank ordering. The Pearson r for this procedure was 0.57. This
correlation coefficient was a reliability estimate for a 15-item test. In order to transform
this split-half reliability into an estimate for the entire test, the Spearman-Brown
prophecy formula was used. The estimated reliability of the entire test was found to
be 0.72.

Treatment
The unit under study was entitled “Conservation of Energy and Momentum” and
included the major concepts work, power, energy, and momentum. The instructor’s
concept map for the unit appears in Appendix B. The unit corresponded to Chapter 6
of Modern Physics by Williams, Trinklein, and Metcalfe (1984), and was the sixth
of seven units in the first semester of secondary school physics.
All students were given a study guide which included a one paragraph introduction;
the 15 objectives of the unit; a list of major concepts; and enabling activities which
includes reading and written assignments such as textbook questions and problems,
conceptual questions, a collisions worksheet, and three experiments-two of which
were required. This procedure was normally followed throughout the course.
The students were used to an informal classroom environment in which they
worked at their own pace. Most students formed partnerships and worked in small
groups. The teacher supplied an introductory motivational lecture, minilectures to
small groups or the whole class (as needed as the unit progressed), and a review at
the end of the unit.
Four of the classes had six weeks of instruction in concept mapping. The instruction
included the submission of concept maps on reading passages on scientific laws and
theories and on the general content of physics. In addition, those four classes submitted
concept maps at the conclusion of the two previous units: forces, friction and torque,
and two-dimensional motion.
Three of the classes (one from each treatment group) took a pretest identical to
the posttest on the first day of the unit. The classes were randomly selected and
randomly assigned as to treatment and pretest conditions. The teacher was the same
for all groups.
320 PANKRATIUS

Control Group. The control group (two classes) followed normal instruction as
delineated above. This group received no concept-mapping instruction and was required
to complete the third experiment (on conservation of momentum and energy) to compensate
for not having to submit concept maps for the unit. The additional experiment (in lieu
of concept maps) was in agreement with instructional practices for the previous two
units.
Post Instructional Mapping Group. These two classes (the low-level treatment
group) were required to submit concept maps at the end of the unit. The teacher
responded to individual questions concerning the structure of students’ concept maps,
but there was no attempt made at large-scale classroom instruction on the concept map
for the unit. This followed the practice for the previous two units.
PreIPost Instructional Mapping Group. Two of the four concept-mapping classes
were required to complete a concept map of the key concepts of the unit under study
prior to instruction, the first or second day of the unit. These two classes (the high-
level treatment group) were encouraged to revise their concept maps as the unit progressed.
Again, the teacher responded to individual questions Concerning the structure of students’
concept maps, but there was no attempt made at large-scale classroom instruction on
the concept map for the unit.

Research Design
The experimental design followed in this study was based on the Solomon four-
group design. The classes were chosen and assigned to one of six experimental groups
at random. The design of the study, along with the distribution of classes and the
number of students assigned to each group, is depicted in Figure 1. 01, 03, and 05
were the pretests. 02, 04, 06, 07, 08, and 09 were the posttests (identical to the
pretests). X1 was the post instructional mapping group. X2 was the prelpost instructional
mapping group.
This design cannot be properly labeled a true experimental design since the individual
subjects were not randomly assigned to experimental and control treatments; however,
it did give greater control over the selection threat to validity than the usual quasiex-
perimental design. The selection threat to validity was diminished by the random
assignment of students to classes at the beginning of the semester, and further, by the
random selection and assignment of intact classes to experimental groups.
BUILDING AN ORGANIZED KNOWLEDGE BASE 321

TABLE 1
Posttest Means

Group Pretest Treatment -


n Mean so
-
-
1 10.0 Control 14 14.3 4.3
2 10.5 x1 14 15.1 1.9
3 7.1 x2 14 14.9 3.3
4 no Control 15 12.6 4.0
5 no x1 15 13.7 3.9
6 no x2 15 17.7 3.6

Results
The group posttest means are reported in Table I. The pretest scores are reported
for the three groups that were exposed to it.
A two-way (factorial) ANOVA was performed to determine if a pretest-posttest
effect was present. One factor was the level of treatment. The other factor was the
pretest condition-either yes or no. The data arrangement for this two-way ANOVA
is depicted in Figure 2. Groups of unequal subjects required the use of a harmonic
mean sample size.
There was no significant interaction effect (see Table 11); thus there was no threat
to internal validity as a result of pretest sensitization. Indeed, the group scoring highest
on the posttest was not exposed to the pretest. The posttest means for the combined
groups were then analyzed.
The statistical test for H(0)1, the first null hypothesis, was a one-way analysis
of covariance (ANCOVA) of the posttest means. The math subscore of the Scholastic
Achievement Test (SATM) was the covariant. The sample sizes were equal for this
analysis. The adjusted and covariant means are reported in Table 111. The correlation
coefficient was 0.50.
The significant omnibus F resulted in the rejection of H(0)l (see Table IV); but
it did not reveal which of the treatments was more significant. A significantF demanded
a further analysis of the data. The next question was, “Which differences were significant?”
In order to accept or reject H(R)I,the first research hypothesis, a single-df
comparison of the adjusted treatment means was accomplished. This comparison (Comp
1) was between the control group and the average of the means of the two experimental
groups (see Table V). The significant F resulted in the acceptance of H(R)l.

~-

Fig. 2. Two-way ANOVA to compare pretest-posttest interaction.


322 PANKRATIUS

TABLE I1
Two-way ANOVA Comparing Pretest-Posttest Effects

Source ss -
df MS F
- sig

Treatment 123.14 2 61.07 4.47 .025


Pretest .22 1 .22 .02 ---
Interaction 71.49 2 35.74 2.62 -_-
Within 1106.67 81 13.66
Total 1300.52 86

TABLE I11
Posttest and SATM Means

Group -n Mean -
SD A d j u s t e d Mean SATM

Control 29 13.4 4.2 13.6 61.6


x1 29 14.4 3.2 14.5 62.3
x2 29 16.4 3.7 16.1 64.0

TABLE IV
One-way ANCOVA of Treatment Means

Source
"ad j -
df M S a dj
-F sig

Between Groups 87.88 2 43.94 4.21 .025


W i t h i n Groups 867.10 83 10.45
Total 954.98 85

TABLE V
ANOVA Comparison of Control and Adjusted Treatment Means

Source ss -
df MS -F sig

Treatment (92.39) (;I


Comp 1 49.49 49~49 4.70 .05
Comp 2 43.90 1 43.90 4.17 .05
W i t h i n Groups 873.16 83 10.52
Total 965.55 85
BUILDING AN ORGANIZED KNOWLEDGE BASE 323

In order to accept or reject H(0)2, the second null hypothesis, a single-dfcomparison


between the two experimental groups was accomplished. This was Comp 2 in Table
V. The significant F resulted in the rejection of H(0)2 and the acceptance of the second
research hypothesis, H(R)2.

Student Concept Maps


Concept maps constructed by the students demonstrated their understanding of
the concept of energy. The preinstructional concept maps (drawn by different students)
shown in Appendix C indicated that both students felt that energy was (or could be)
caused by collisions. One student stressed that energy was saved by conservation.
Both concept maps evidenced confusion with the concept power. Gravitationalpotential
energy was misunderstood by one student. The two concept maps shown in Appendix
D and Appendix E demonstrated the development of a student’s understanding as the
unit progressed. Both of these concept maps were correct as to relationships between
concepts, but the second one (post instruction) was more fully developed. The concept
map shown in Appendix F exhibited the elaborate structure and wealth of equations
inherent in the unit under study. The concept map shown in Appendix G displayed
the individuality of a student who stressed the conservation laws. The concept map
in Appendix E clearly showed the relationship between momentum and change in
impulse. That relationship was not implicitly shown in the student concept maps in
Appendices F and G.

Conclusions

It can be concluded that for this sample (upper-middle-class high school physics
students) groups that received instruction on concept mapping scored significantly
higher than groups that received standard instruction. The level of treatment (amount
of concept mapping) was related to the gain in achievement. Groups that mapped
concepts prior to, during, and subsequent to instruction recorded higher posttest scores.

Discussion
The principle finding of statistical significance between the treatment means and
the control group, as well as between the two treatment groups, was encouraging. The
group that mapped concepts prior to and subsequent to instruction (high-level treatment
group) had posttest scores 18.4% higher than the control group and 11.O% higher than
the group that mapped concepts at the end of instruction (low-level treatment group).
The posttest scores of the latter group were only 6.6% higher than the control group.
The degree of concept mapping was clearly the major factor.

Limitations
A serious threat to the validity of any experimental study was low statistical power.
The power of this experiment was estimated before the study to be 0.91. Smaller
differences between the treatment means and a smaller variance combined to lower
the estimated post experiment power to .78. Although this value appeared to indicate
324 PANKRATIUS

that the statistical sensitivity of the experiment was close to the lower limit, the power
of the study was sufficient to detect statistically significant differences.
Two related threats to the validity of this study centered on crosstalk among the
subjects and the response of the subjects. The students were in six intact classes with
the same instructor (this investigator). They talked among themselves and were well
aware that a “study” was in progress. The third group that took the pretest apparently
realized that it did not count; hence, they did not try too hard. Their pretest score was
considerably lower than the other two pretest groups.
The last period of the day was the posttest-only group that received the high-level
treatment. They scored higher than any other group on the posttest. While some
discussion of test questions was possible, the other two afternoon classes (one was a
control group, the other a low-level treatment group) scored lower on the posttest than
the morning classes. It was felt that the SATM covariant accounted for any differences
among the student population; yet there was an elusive missing factor. Seniors in the
school district of the study were permitted to leave campus after five or six periods if
they were on track for graduation. Those that remained for seven periods were often
involved in extracurricular activities and tended to be highly motivated. This factor
was not accounted for in the study.
A minor threat to the study was the high mortality rate (40%). Scheduling of the
posttest on the day before winter vacation led to major conflicts for many students.
A winter carnival ball on the night before the posttest kept over 30 students out of
class on the day of the posttest.
Survey of the physics students in the course indicated that the 87 students involved
in the study did not differ substantially from the 145 students enrolled in the course.
However, the ability to generalize the findings to other high school physics students
was limited.
It cannot be overlooked that those students in the treatment groups spent more
time on task than those in the control groups. An additional experiment was assigned
to the control groups in an attempt to control for this time on task factor.

Implications for Educational Practice


In spite of the threats to experimental validity, concept mapping has a positive
effect on achievement. At the same time the mapping of concepts did not diminish
the content studied in the course. In fact, mapping of the whole course (by the instructor
subsequent to this study) appears to have had the desirable effect of linking seemingly
disparate units. In addition, concept mapping by the students can convince them to
examine the continuity of subject matter.
Class time is required to teach concept mapping. The eight-week period of this
study may seem to be a short time frame for students to master concept mapping.
Novak (personal communication, December, 1986)recommends six months; however,
this experiment did involve above-average students who appeared to have mastered
concept mapping prior to the unit under study. The time requirement to teach concept
mapping is well worth the benefits of increased achievement and better understanding.

Recommendations
This study answers the question regarding the relationship between concept mapping
and achievement in a positive fashion. It also raises several related questions:
BUILDING AN ORGANIZED KNOWLEDGE BASE 325

1. What is the relationship between concept mapping and student learning


styles? Some students are highly resistant to concept mapping-they feel that the
concept map adds little to their understanding. Some poor mappers with low
achievement scores tend to construct linear maps with limited branching.
2. What is the relationship between the quality of the concept map and achieve-
ment? A previous study by this author (Pankratius, 1986)reveals a weak correlation
between concept map scores and achievement. Although age may be a factor in
the 1986 study, the brightest students still seem to draw the poorest maps.
3. What is the relationship between concept mapping and student misconceptions?
Are misconceptions clarified by the process of concept mapping?
4. What is the relationship between concept mapping and the processes of
science? Is concept mapping another way of learning science?

Concept mapping is a powerful teaching strategy that requires students to participate


in the building of their knowledge bases. The achievement gains seen in this study
can help students to make the transition from rote to meaningful learning.

References

Ault, R.A. Jr. (1985). Concept mapping as a study strategy in earth science.
Journal of College Science Teaching, 15, 38-44.
Ausubel, D.P., Novak, J.D., & Hanesian, H. (1978). Educational Psychology:
A Cognitive View, 2nd ed. New York, NY: Werbel & Peck.
Bouquet, W.S. (1982). An applicationof Ausubel’s learning theory to environmental
education: a study of concept mapping in a college natural resources management
course. (Doctoral dissertation, Ohio State University, 1982.) Dissertation Abstracts
International, 43, 1515A (University Microfilms No. DA 8222053).
Campbell, D.T. & Stanley, J.C. (1963). Experimental and Quasi-Experimental
Designs for Research. Chicago: Rand McNally and Company.
Champagne, A.B., Klopfer, L.E., Desena, A.T., & Squires, D.A. (1981). Structural
representations of student’s knowledge before and after science instruction. Journal
of Research in Science Teaching, 18(2), 97- 112.
Chi, M.T.H., Rltovich, P.J., & Glaser, R. (1981). Categorizationand representation
of physics problems by experts and novices. Cognitive Science, 5, 121- 152.
‘ Clibum, J.W., Jr. (1986). Using concept maps to sequence instructionalmaterials.
Journal of College Science Teaching, 14(5), 38-44.
Eylon, B.S. & Reif, F. (1979, April). Effects of internal knowledge organization
on task performance. Paper presented at the annual meeting of the American Educational
Research Association (ERIC Document Reproduction Service No. ED 171 804).
Eylon, B.S. & Reif, F. (1984). Effects of knowledge organization on task per-
formance. Cognition and Instruction, 1(l), 5-44.
Finkel, E. (1983). Three approaches to group teaching of verbal problem solving
in secondary school physics. (Doctoral dissertation, Georgia State University, 1983.)
Dissertation Abstracts International, 44, 720A.
Fraser, K. & Edwards, J. (1985). The effects of training in concept mapping on
student achievement in traditional tests. Research in Science Education, 15, 158- 165.
Proceedings of the Annual Conference of the Australian Science Education Research
Association (ERIC Document Reproduction Service No. ED 267 974).
326 PANKRATIUS

Greeno, J.G. (1980). Trends in the theory of knowledge for problem solving. In
D.T. Tuma & F. Reif, Eds., Problem Solving and Education: Issues in Teaching and
Research. Hillsdale, NJ: Lawrence Erlbaum Associates, pp. 9-23.
Keppel, G. (1982). Design and Analysis: A Researcher’s Handbook. Englewood
Cliffs, NJ: Prentice-Hall.
Larkin, J.H. & Reif, F. (1979). Understanding and teaching problem solving in
physics. European Journal of Science Education, 1(2), 191-203.
Lehman, J.D., Carter, C., & Kahle, J.B. (1985). Concept mapping, vee mapping,
and achievement: results of a field study with black high school students. Journal of
Research in Science Teaching, 22(7), 663-673.
Lin, H. (1982). Learning physics vs. passing courses. The Physics Teacher, 20(3),
151-157.
Moreira, M.A. (1977). An Ausubelian approach to physics instruction: An experiment
in an introductory course in electromagnetism. (Doctoral dissertation, Cornell University,
1977 .) Dissertation Abstracts International, 38, 5378A (University Microfilms No.
780 164 1).
Moreira, M.A. (1979). Concept maps as tools for teaching. Journal of College
Science Teaching, 8(5), 283-286.
Novak, J.D. (1976). Understanding the learning process and effectiveness of
teaching methods in the classroom, laboratory, and field. Science Education, 60(4),
493 - 5 12.
Novak, J.D. (1977a). A Theory of Education. Ithaca, NY: Cornell University
Press.
Novak, J.D. (1977b). An alternative to Piagetian psychology for science and
mathematics education. Science Education, 61(4), 453-477.
Novak, J.D. (1981). Applying learning psychology and philosophy of science to
biology teaching. The American Biology Teacher, 43( I), 12-20.
Novak, J.D. (1984). Application of advances in learning theory and philosophy
of science to the improvement of chemistry teaching. Journal of Chemical Education,
61(7), 607-613.
Novak, J.D. & Gowin, D.B. (1984). Learning How to Learn. New York, NY:
Cambridge University Press.
Novak, J.D., Gowin, D.B., & Johansen, G.T. (1983). The use of concept mapping
and knowledge vee mapping with junior high school students. Science Education, 67,
625-645.
Ontario Institute for Studies in Education (198 1) Ontario Assessment Instrument
PooE: Physics-Senior Division (OAIPPSD) (ON 01650). (Available from OISE Pub-
lications, 252 Bloor Street West, Toronto, Ontario, Canada, M5S 1V6.)
Pankratius, W.J. (1986). Concept mapping and chapter outlining in support of
meaningful learning in a middle level ninth grade physical science course. Paper
presented at the annual meeting of the Georgia Academy of Sciences, Milledgeville,
GA, May, 1986.
Richardson, J.J. (1981). Problem solving instruction for physics. (Doctoral dis-
sertation, University of Colorado, 198 1 .) Disserration Abstracts International, 42,
3536A (University Microfilms No. AAD-82-00820).
Stewart, J., Van Kirk, J., & Rowell, R. (1979). Concept maps: a tool for use in
biology teaching. The American Biology Teacher, 41(3), 171 - 175.
True, D.S. (1975). Problem solving instruction in physics. (Doctoral dissertation,
BUILDING AN ORGANIZED KNOWLEDGE BASE 327

Columbia University, 1974.) Dissertation Abstracts International, 36, 6532A (University


Microfilms NO. AAD-75-7856).
Williams, J.E., Trinklein, F.E., & Metcalfe, H.C. (1984). Modern Physics. New
York: Holt, Rinehart, & Winston.

Manuscript accepted June 23, 1989


328 PANKRATIUS

is
transferred

Appendix B. Instructor’s concept map for the unit under study.


BUILDING AN ORGANIZED KNOWLEDGE BASE 329

measured by

Appendix C. Students’ preinstruction concept map high-level treatment group.


0
0
0

measLred in

rneasLred in

Appendix D. Reinstruction concept map high-level treatment group.


W
2
P

equ tion

P -'wn

Appendix E. Postinstruction concept map high-level treatment group.


-.
w
w
h)

Appendix F. Postinstruction concept map high-level treatment group.


9
z

Appendix G . Postinstruction concept map high-level treatment group.

You might also like