Professional Documents
Culture Documents
Accepted Manuscript: Computers & Education
Accepted Manuscript: Computers & Education
PII: S0360-1315(18)30168-4
DOI: 10.1016/j.compedu.2018.06.025
Reference: CAE 3389
Please cite this article as: Yang T.-C., Chen M.C. & Chen S.Y., The influences of self-regulated learning
support and prior knowledge on improving learning performance, Computers & Education (2018), doi:
10.1016/j.compedu.2018.06.025.
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to
our customers we are providing this early version of the manuscript. The manuscript will undergo
copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please
note that during the production process errors may be discovered which could affect the content, and all
legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT
The influences of self-regulated learning support and prior knowledge on
improving learning performance
Tzu-Chi Yang
Institute of Information Science, Academia Sinica, 128, Academia Road, Section 2,
Nankang, Taipei 115, Taiwan
E-mail: tcyang.academic@gmail.com
PT
Meng Chang Chen
Institute of Information Science, Academia Sinica, 128, Academia Road, Section 2,
Nankang, Taipei 115, Taiwan
RI
E-mail: mcc@iis.sinica.edu.tw
Sherry Y. Chen*
SC
Graduate Institute of Network Learning Technology, National Central University, 300,
Jung-da Rd., JhongLi 320, Taiwan
E-mail: sherry@cl.ncu.edu.tw
U
*Corresponding author
AN
Abstract
M
this end, this study aimed to examine how high prior knowledge students (HPKs) and
low prior knowledge students (LPKs) behaved differently when interacting with a
TE
learning process. Moreover, the LPKs and HPKs behaved similarly in the forethought
and self-reflection phases but some behavior differences were found in the
C
performance phase, where the LPKs relied on the notes and sought support the
dashboard and quiz records while the HPKs did not demonstrated such a tendency.
AC
Our results’ theoretical and methodological implications and possible applications for
further research are also discussed.
ACCEPTED MANUSCRIPT
Highlights
High and low prior knowledge students used different SRL strategies.
Prior knowledge has effects on students’ SRL behaviors in the performance
phase.
The SRL support system was helpful to remove the gap of learning performance
caused by prior knowledge.
PT
Keywords: Evaluation of CAL systems; Interactive Learning Environments;
Teaching/learning strategies
RI
U SC
AN
M
D
TE
C EP
AC
ACCEPTED MANUSCRIPT
The Influences of self-regulated learning support and prior knowledge
Abstract
PT
Self-regulated learning (SRL) is helpful to students. On the other hand, prior knowledge has great
effects on students’ self-regulation and learning performance. To this end, this study aimed to
RI
examine how high prior knowledge students (HPKs) and low prior knowledge students (LPKs)
SC
behaved differently when interacting with a SRL environment. To achieve this aim, we proposed a
self-regulated learning support system (SRLSS) for a mathematical course. The results showed that
U
the gap of learning performance between the HPKs and LPKs was removed after a long-term
AN
learning process. Moreover, the LPKs and HPKs behaved similarly in the forethought and self-
reflection phases but some behavior differences were found in the performance phase, where the
M
LPKs relied on the notes and sought support the dashboard and quiz records while the HPKs did not
D
demonstrated such a tendency. Our results’ theoretical and methodological implications and possible
TE
1. Introduction
C
Self-regulated learning (SRL) refers to a scenario, where students are active and responsible for their
AC
own process (Zimmerman, 2008) and are able to be self-aware, knowledgeable and to decide an
approach to learning (Corno, 1986; Zimmerman, 1990). In past decades, a growing body of research
indicated that the SRL greatly affected students’ learning performance and academic achievement
(Wang, 2011). Specifically, strong SRL skills also predict high self-efficiency and satisfaction,
which can result in better learning outcomes (Artino, 2008). Researchers indicated that students with
1
ACCEPTED MANUSCRIPT
strong SRL skills were more likely to be successful either in classrooms (Pintrich, 2000; Bellhäuser,
In other words, the SRL has positive effects on student learning (Fisher, & Ford, 1998). As
suggested by Zimmerman (2000) and Zeidner, Boekarts and Pintrich (2000), students with the SRL
are expected that they (1) can improve their abilities to learning through selecting and using
PT
metacognitive and motivational strategies; (2) can proactively select, structure and even create a
RI
learning environment that is advantageous to themselves; (3) can play a critical role in choosing the
form and amount of instruction they need. However, doing the aforementioned SRL activities well
SC
is not a simple process because there is a need to rely on prior knowledge (Moos & Azevedo, 2008).
This is due to the fact that prior knowledge can affect students’ motivation and learning
U
approaches (Winters, Greene, & Costich, 2008). As shown in Greene, Moos, Azevedo, & Winters,
AN
(2008), high prior knowledge students (HPKs) could demonstrate more self-reflective and
M
monitoring behaviors than lower prior knowledge students (LPKs). Additionally, Bernacki, Byrnes,
& Cromley (2012) indicated that HPKs could usually use a more active learning approach while
D
LPKs passively followed instruction. Subsequently, Song, Kalet and Plass (2016) found that HPKs
TE
usually showed better learning performance than LPKs. These studies suggested that there is a strong
Such a link raises an issue that the SRL may not be very beneficial to LPKs. For example,
C
Winters, Greene and Costich (2008) found that LPKs had limited improvement and less tended to
AC
engage in planning and monitoring learning activities while HPKs might demonstrate better
improvement and high engagement in a SRL context. This might be because HPKs had stronger
(Dunlap, 2005). In other words, metacognitive skills could facilitate student learning in the SRL
context (Winne, 1995; Lehmann, Hähnlein, & Ifenthaler, 2014), where students were requested to set
2
ACCEPTED MANUSCRIPT
goals, plan learning strategies, self-monitoring, and evaluate their strategies. On the other hand,
metacognitive skills were considered as critical skills for planning and selecting learning strategies,
assessing the effectiveness of learning strategies, and changing learning behaviors and strategies
when necessary. Accordingly, metacognitive skills provided proficiencies that the SRL needed so
PT
However, LPKs lack such metacognitive skills (Schunk, 2005; Liu, Andre, & Greenbowe,
RI
2008). This might be the reason why researchers attempted to investigate how to help LPKs to
improve their learning performance as well as SRL skills by providing additional support and
SC
training for LPKs (Kappa, 2001; Cleary, & Zimmerman, 2004; Spruce, & Bol, 2015). For example,
Azevedo, Moos, Greene, Winters, & Cromley (2008) proposed an extended SRL support framework,
U
where agents/tutors were applied to facilitate students to monitor their learning status and seek extra
AN
help. Their results addressed that the learning performance of the LPKs could be significantly
M
enhanced and they demonstrated more advanced learning skills in the SRL context. Recently, results
from other studies also suggested that external support was helpful for improving the learning
D
performance of the LPKs in SRL contexts (e.g., Lehmann, Hähnlein, & Ifenthaler, 2014).
TE
However, these studies focused on specific or single support for SRL and lack
comprehensive understandings of how students’ behaviors were affected by the SRL (Taub, Azevedo,
EP
Bouchet, & Khosravifar, 2014; Trevors, Duffy, & Azevedo, 2014). Accordingly, there is a need to
C
trace and analyze students’ learning behaviors (e.g., Roll, Baker, Aleven, & Koedinger, 2014), which
AC
can not only provide behavioral evidence that reflects the influences of the SRL on student learning,
but also offers researchers a deep insight of the effects of the SRL (Biswas, Jeong, Kinnebrew,
In this vein, the aims of this study are two-fold. One is to propose a SRL environment to
support students. Among various topics, the SRL environment proposed in this study focused on
mathematical learning. Mathematical problem solving relies on regulating the selection of various
3
ACCEPTED MANUSCRIPT
forms of knowledge (Zimmerman, 1989). For example, to solve a mathematical question, student
need to recall what they have learn, to select related solutions (e.g., definition, formula), and to use
the solutions correctly. To achieve the aforementioned activities, student need to aware their
learning status, to deal with their weaknesses, or to change their learning strategies. Such a process
corresponds to the core theme of the SRL. In other words, SRL skill might greatly affect learning
PT
performance in mathematic learning (Fuchs, 2003; Kramarski, & Gutman, 2006). Hence,
RI
mathematics was considered as the subject topic of the SRL environment developed in this study.
The other aim is to identify the effects of the SRL on student learning from a prior knowledge
SC
perspective, including learning performance, learning behaviors and learning perceptions. To
correspond to these two aims, the research questions of this study are:
1.
U
whether the proposed SRL environment is helpful to improve students’ learning
AN
performance?
M
2. whether LPKs and the HPKs behave differently in the proposed SRL environment?
To find the answers to these research questions, we examine the influences of prior
TE
knowledge on SRL across the three phases identified by Zimmerman (2000). By doing so, our study
can not only reveal the importance of understanding the effects of prior knowledge on SRL, but also
EP
provide guidance on how to offer adequate support for HPKs and LPKs in SRL environments.
C
AC
2. System design
Various tools have been applied to provide additional support in existing SRL environments,
including test/quiz (e.g., Ekholm, Zumbrunn, & Conklin, 2015), note-taking (Kauffman, 2004) and
feedback (Sung, Liao, Chang, Chen, & Chang, 2016). However, such SRL environments mainly
provided a single tool. As suggested by some researchers (Biswas, Jeong, Kinnebrew, Sulcer, &
Roscoe, 2010; Clark, 2012), providing learners with multiple tools can engage them during the SRL
4
ACCEPTED MANUSCRIPT
process. To this end, this study proposed a self-regulated learning support system (SRLSS), which
integrated various tools to facilitate students to learn mathematics during the SRL process.
The design rationale of the SRLSS was based on the Cleary and Zimmerman’s SRL cycle
(2004), where learning behaviors were divided into three phase: forethought phase, performance
phase and self-reflection phase. To correspond to these three phases, the SRLSS consisted of three
PT
modules, i.e., forethought module, performance module, self-reflection module, each of which
RI
included various tools that support students to develop SRL skills. The details of each module are
U SC
AN
M
D
TE
C EP
In the forethought phase, students analyze their tasks, set goals and plan strategies to accomplish
their goals (Zimmerman, 2002). Thus, the FM (Fig. 2) provides students with tools, which can
facilitate them to set goals, manage time and plan strategies. Furthermore, each student had a
5
ACCEPTED MANUSCRIPT
database, which recorded how they achieved these activities so that they could check their current
● To set goals: Tools were provided to facilitate students in externalizing their goals.
Moreover, the expected scores of examinations and the anticipated correctness rate of
PT
quizzes were automatically presented so that students could be motivated to achieve such
RI
goals.
● To manage time: There were also tools, which could facilitate students in managing their
SC
time, including time spent on studying, taking quiz and reviewing learning materials.
● To plan strategies: Tools that could facilitate students in planning their strategies were also
U
available. Further to their own strategies, some strategies suggested by teachers were also
AN
included so that they could be guided to create strategies for themselves.
M
D
TE
C EP
AC
6
ACCEPTED MANUSCRIPT
In general, self-monitoring and self-control were major strategies that student used during the
performance phase (Zimmerman & Kitsantas, 2007). Thus, students needed to implement learning
strategies and to correspond their performance to their goals to determine progress in this phase. To
PT
help students achieve their goals effectively, various online quizzes were provided (Fig. 3). Such
RI
quizzes included a standardized quiz, which covered the fixed scope of the subject content, i.e., the
weekly quiz and summarized quiz, and a customized quiz, which allowed students to choose the
SC
scope based on their own needs, i.e., unit quiz and personal quiz. The details are listed below.
In addition to the online quizzes, the SRLSS provided the following tools based on
TE
● To review learning portfolio: Tools were provided to facilitate students in reviewing their
EP
learning portfolios, such as quiz history and related topics, highlighted questions, correct and
C
incorrect answers to questions. Student could monitor the performance and identify their
AC
● Personal note: Students were allowed to make personal notes anytime. Moreover, they could
add questions, solutions and comments to their own personal notes when taking the quizzes
and reviewing quiz results so that students could review their questions by browsing the
personal notes.
● To seek help: Text-based or video-based solutions were provided for students. The text-based
7
ACCEPTED MANUSCRIPT
solution presents detailed solutions while the video-based solution provides step-by-step
guidance and teachers’ explanations for the solution. By doing so, student could deal with
their problems or improve their understandings by choosing a suitable solution based on their
preferences.
PT
RI
U SC
AN
M
D
TE
During the self-reflection phase, students reflected and evaluated how they achieved goals according
to their performance. Students assessed their successes or failures, adjusted their self-efficacy and
reflect themselves whether they had accomplished their goals so that they could adjust their learning
strategies or goals for next steps. To facilitate the aforementioned reflection and evaluation, the SRM
● Dashboard: The dashboard provided various information related to each student’s learning
8
ACCEPTED MANUSCRIPT
status (Fig. 4). Specifically, the history of a student’s activities on the SRLSS was analyzed and
visualized in the dashboard, where student could see rich information, such as the correctness
rate of taking the quizzes and the average correctness rate of other classmates, performance on
each topic, the reports of how they used the SRLSS. Moreover, some suggestions were also
PT
● Self-reflection Tools: There were tools (Fig. 5), which presented the learning goals and strategies
RI
planned by the students so that they could aware their performance and compared with expected
performance.
SC
● Self-assessment Tools: Tools were provided to encourage students to assess and evaluate
themselves, including satisfaction, attribution, adaptations, etc. Typical questions were “how
U
satisfied are you with your performance until now”, “what is distracted your study” and “what do
AN
you need to do to improve your performance on your next steps”.
M
D
TE
C EP
AC
9
ACCEPTED MANUSCRIPT
PT
RI
U SC
Fig. 5. An example of prompt personal goals and self-assessment tool.
AN
M
In brief, the SRLSS is theoretically grounded within existing research in the area of SRL. To
have a deep understanding of the influences of the SRLSS on student learning, an experiment was
D
conducted to answer our research questions described in Section 1. The following section describes
TE
the details of our experimental design, including participants, procedures and measures used in the
experiment.
C EP
AC
3. Experimental design
3.1. Participants
The participants of this study were 60 students, who studied in a high school in north Taiwan. All of
the students were taught by the same instructor who had taught them for at least one year. Prior to
conducting the experiment, they had no experience in using such a SRL support system.
10
ACCEPTED MANUSCRIPT
3.2. Instruments
3.2.1. Pre-Test
The pre-test was used to identify participants’ prior knowledge of the basic mathematics in high
schools (K9 to K10), including the understandings of the subject content of a specific concept and
the processes and procedures related to the concept. For instance, student not only need to describe
PT
the definition of De Morgan’s laws, but also they were requested to use the laws to proof some
RI
equations. The pre-test consisted of five multiple-choice-question, five semi-structured question, and
five calculation or proof questions. Table 1 gives an example of these three types of questions. The
SC
perfect score of the pre-test was 100.
Table 1
U
AN
An example of pre-test.
Multiple-choice-question Given A, B, C sets, A' is the complementary set of A. Please select all
correct statements: (1) A∪B=B ⇒ A⊂B (2)A⊂B ⇒ A-B=∅ (3) A⊂C
M
and B⊂C ⇒ (A∪B)⊂C (4) A⊂B ⇒A'⊂B' (5) (A⊂C) ⊂(B∩C) ⇒ A⊂B.
Calculation or proof questions Given A={1,2,3}, B={2,3,4}, and U={1,2,3,4,5}. Assume that A and B
are subsets of a universe U. Proof that (1) (A∩B)' =A'∩B' (2)
TE
(A∪B)'=A'∪B'
EP
There were four formative tests, two of which covered specific topics based on their course progress.
C
The other two tests were the post-test and delayed test, which were used to check students’ learning
AC
performance during the learning process (Guskey, 1980). The post-test was used to assess the
learning performance after they interacted with the SRLSS while the delayed test was used to
measure the long-term persistence of the influences of learning with such a SRL environment. Both
the post-test and the delayed test covered all relevant topics of the teaching unit and were
11
ACCEPTED MANUSCRIPT
3.2.3. Questionnaire
each student at the end of the learning process. The design rationale of this questionnaire was to
examine (a) whether the students agreed that the SRLSS could help them learn the mathematical
course and (b) how the students reacted to the SRLSS. Accordingly, the participants were requested
PT
to answer the following questions
RI
(A) Do you think the SRLSS was beneficial to you, in terms of learning the mathematics course?
If so, please give three sentences to describe your experience. If not, please also give your
SC
reasons.
(B) Does the SRLSS provide you with sufficient tools for improving your learning performance?
U
(C) Please depict three tools that were the most helpful for you and explain why such tools were
AN
helpful,
M
(D) Please give your suggestions for improving the next version of the SRLSS.
D
3.3. Procedure
TE
The overall procedure of this study is described in Fig. 6. In the first stage, students received face-to-
face instruction for four weeks. Furthermore, students needed to take the pre-test to identify their
EP
prior knowledge. The participants whose pre-test scores were lower than the average, were assigned
C
to a low prior knowledge group (LPK). Otherwise, the participants, whose pre-test scores were
AC
higher than the average, were allocated to a high prior knowledge group (HPK). In other words, the
participants were divided into two groups based on their scores from a pre-test.
In the second stage, the students were then asked to complete weekly assignments and
quizzes on the SRLSS after each lecture was delivered. In other words, they would generally interact
with the SRLSS at least once each week. Meanwhile, how students interacted with the SRLSS was
recorded in a database so that their learning behaviors could be analyzed. During the learning
12
ACCEPTED MANUSCRIPT
process, they were free to access resources (e.g., note, solution, other tools) provided by the SRLSS.
Additionally, they needed to take the formative tests every two weeks and they were also
administered with the post-test. In the end, they needed to fill out the questionnaire so that their
perceptions could be identified. After three weeks, they were requested to take the delayed test so
PT
RI
U SC
AN
M
D
TE
EP
The independent variable of this study was the levels of prior knowledge that students possessed. In
order to get a deep understanding of the SRLSS, we collected comprehensive information so three
dependent variables were included, i.e., learning performance, learning behaviors and learning
perceptions toward the SRLSS. Regarding learning performance, the scores from formative tests,
post-test, delayed test were applied to identify if there were statistically significant differences
13
ACCEPTED MANUSCRIPT
between HPKs and LPKs. Furthermore, a significance level of p < 0.05 was adopted for the study.
Regarding learning behaviors, the lag sequential analysis approach (LSA) (Bakeman & Gottman
1997), which is an approach that examines whether particular sequences of behavior occur more or
less frequently than what would be expected, was applied to identify learning behavior of HPKs and
LPKs. In addition, qualitative data from students’ responses to the questionnaire were used to
PT
identify their perceptions, which were applied to illuminate the findings from learning performance
RI
and learning behavior.
SC
4. Results
U
AN
The result (Table 2) showed that the LPKs obtained lower scores than the HPKs in formative
test 1, formative test 2 and the post-test. This result is reasonable because prior knowledge has
M
positive effects on learning performance (Dochy, Segers, & Buehl, 1999). However, no significant
difference was found for the delayed test scores. A difference between the post-test and the delayed
D
test lied within the fact that the former took place immediately after students interacted with the
TE
SRLSS whereas the latter was held long after the SRLSS was administered. Thus, this finding
suggested that the LPKs could still benefit from the SRLSS though they might need longer time.
EP
Furthermore, the paired t-test was applied to examine score differences between each test
C
(Table 3). Regarding the HPKs, there were significant differences between the scores of formative
AC
test 1, formative test 2 and the post-test and the score of the pre-test while no significant difference
existed between the delayed test score and pre-test score. Regarding the LPKs, there was a
significant difference between the post-test score and the pre-test score and between the delayed test
score and pre-test score while no significant differences were found between the remaining test
scores (i.e. formative test 1 and formative test 2) and the pre-test score. The results showed that the
learning performance of the LPKs was significantly improved on post-test and delayed test, when
14
ACCEPTED MANUSCRIPT
compare with the pre-test. In other words, the SRLSS could enhance the LPKs performance and
remove the gap caused by prior knowledge (Kapa, 2001; Mitchell, Chen, & Macredie, 2005) from a
long-term perspective.
Table 2
The t-test result of the learning performance of the two groups.
PT
Group N Mean S.D. p
RI
HPK 29 74.31 20.06
SC
HPK 29 57.21 25.22
U
HPK 29 53.04 20383
AN
Post-test LPK 31 51.68 24.92 .027*
Table 3
The paired t-test result of the LPKs’ learning performance.
EP
Mean S.D. p
C
15
ACCEPTED MANUSCRIPT
4.2. Students’ behaviors on the SRLSS
To analyze the behaviors demonstrated by students, we firstly coded students’ behaviors and
calculated the obtained frequencies of each code; after doing so, the LSA was used to uncover their
learning behaviors. The individual behaviors of the participants were coded using the coding scheme,
which included 19 codes related to the three phases of the SRL model (Table 4). Thereby, the
PT
frequency of each code was presented in Fig. 7.
RI
Table 4
The code schema for general behaviors.
SC
Behavior Codes Description
Performance phase
U
Login I To log in the SRLSS
Take a personal quiz S To take a quiz which covered topics what they chosen
M
Take a summarized quiz M To take a quiz that includes every topic they learnt.
D
To show the latest quiz result R To click to show the outcome of the latest quiz
TE
To show previous quiz outcome C To select to show an outcome of the completed quizzes
EP
To show text solution A To click the button to show the answers to the questions
16
ACCEPTED MANUSCRIPT
strategies
PT
In order to have an overview of learning behaviors of HPKs and LPKs, the frequency of each
behavior obtained was calculated so that the distribution of behaviors of these two groups could be
RI
uncovered (Fig 7). For example, the LPKs took the weekly quiz (W) 821 times while the HPKs took
SC
it 269 times based on the frequencies of obtained codes (Fig. 7). As shown in Fig. 7, code A (open
text solution) has the highest frequencies for both groups. This finding suggested that both groups
U
were most concerned with the text solutions.
AN
However, the LPKs differed from the HPKs in behaviors with the second and third high
frequencies. Regarding the second-high frequencies, the behavior for the LPKs was to take the
M
weekly quiz (W) while that for the HPKs was to watch the video solutions (V). Regarding the third
D
high frequencies, the behavior for the LPKs was to watch the video solutions (V) while that for the
TE
HPKs was to plan learning strategies (SP). This finding implied that the LPKs paid attention to the
quiz while the HPKs emphasized on their own learning strategies. Such a difference may be
EP
reasonable because the LPKs usually rely on resources provided by teachers (Liu, Andre, &
Greenbowe, 2008) while the HPKs usually develop their own learning strategies to accomplish their
C
learning goals (Shen, & Chen, 2006). Accordingly, there is a need to investigate how the LPKs and
AC
HPKs behaved differently in the SRLSS. To this end, we conduct the LSA for their learning
17
ACCEPTED MANUSCRIPT
PT
RI
Fig. 7. The frequencies of obtained codes.
SC
4.4. Lag Sequential analysis
U
According to the rationale of the LSA, participants’ behaviors were coded in the
AN
chronological order of their occurrences. For example, after logging into the SRLSS (I), a student
takes a weekly quiz (W), then the student clicks the button to show the dashboard for his/her learning
M
status (D), and take a unit quiz (U) then add a note (N); this series of behaviors was thus coded as I
W D U N. Based on this rationale, the adjusted residuals tables of behavior sequence for the LPKs
D
and HPKs were conducted (Table 5 and Table 6). The first column presents the starting behaviors
TE
and the first row describes the behaviors that immediately occur after the starting behaviors finish.
EP
The numbers represent the z-score of a column behavior occurred immediately after a row behavior
ended. The behavior sequence would be recognized as a significant behavior sequence if the z-score
C
For instance, we can find that the z-score of row I, column W were 16.06 and 10.44 for LPKs
and HPKs, respectively. Thus, an argument can be made that both LPKs and HPKs tended to take
the weekly quiz after logging the system (I→W). The behavior analyses were divided into two parts.
One part was related to behavior shown in the performance phase (Table 5 and Table 6) while the
other part was concerned with behavior demonstrated in the forethought and self-reflection phases
(Table 7 and Table 8). We used such an approach because students’ behaviors in the forethought
18
ACCEPTED MANUSCRIPT
phase were significantly inter-correlated with those in the self-reflection phase while behaviors
demonstrated in the performance phase might be the outcomes of students’ forethought and self-
reflection processes (Cleary, & Zimmerman, 2001; Goetz, Preckel, Pekrun, & Hall, 2007).
Table 5
PT
The result of the sequential analysis of behavior demonstrated by the LPKs.
Given/
I W U S M A V D R C N P
Target
RI
I 0.94 16.06* 7.9* -1.38 6.85* -14.29 -10.32 8.06* 6.5* -5.09 -7.24 -3.07
W -1.02 5.44* 6.5* 0.74 -1.42 0.31 -6.17 -0.38 3.37* -4.68 2.12* -5.53
SC
U 4.82* -2.8 6.76* -2.05 -0.53 1.07 -3.91 -1.31 -0.89 -3.1 4.14* -3.81
S 1.69 -1.82 -2.49 11.52* -1.39 2.16* -1.41 -2 2.21* -1.58 0.16 -1.74
R -0.74 0.05 -2.57 0.62 -1.41 0.34 0.57 -1.19 2.22* 11.12* -3.74 -1.99
M
C -1.59 -4.24 -2.79 -0.25 -2.06 0.09 -2.3 -3.07 -1.19 36.25* -3 -1.76
A -3.08 -9.33 -6.73 -0.42 -3.79 20.62* 5.24* -1.53 -5.29 -1.87 -1.52 -6.72
D
V 1.67 -7.47 -4.89 -2.12 -3.07 -3.58 25.02* -0.47 -5.49 -1.56 -4.62 -4.35
TE
N -3.5 -7.1 -4.6 -2.24 -2.56 6.88* 0.19 -3.34 -2.42 -3.33 21.34* 0.23
P -1.57 -2.93 -3.53 -1.74 0.06 -7.26 -5.25 -4.34 -0.67 -2.59 -3.68 53.45*
C EP
AC
19
ACCEPTED MANUSCRIPT
Table 6
The result of the sequential analysis of behavior demonstrated by the HPKs.
Given/
I W U S M A V D R C N P
Target
I 0.99 10.44* 12.54* 6.3* 0.48 -12.43 -7.6 5.97* 6.74* -4.03 -5.55 -3.9
W -0.39 6.17* 9.38* -1.63 3.77* -3.61 -3.69 -2.88 5.85* -2.93 -0.68 -4.15
U 5.93* -1.79 -0.93 -3.75 1.24 4.19* -3.28 0.28 0.95 -3.21 1.6 -4.01
PT
S 1.19 -2.24 -3.18 4.14* -1.11 -0.83 -1.43 3.8* 0.91 -2.04 5.52* -2.67
M 2.9* 2.27* -0.17 -1.11 6.63* -1.85 -1.7 -1.42 0.59 -0.97 0.47 -0.48
RI
A -2.63 -5.59 -6.53 -4.28 -2.32 20.65* 7.76* -3.22 -6.48 -3 -4.96 -8.09
18.36*
SC
V -1.85 -2.24 -4.65 -1.43 -1.05 -0.05 -2.63 -2.98 -3.52 -1.8 -4.68
D 6.38* -0.96 5.07* 8.76* 1.64 -8.6 -5.63 9.22* 1.27 -2.84 -4.11 -2.95
R -1.21 3.85* -0.91 0.55 -1.19 -2.72 -1.98 -1.05 4.45* 12.87* -3.13 -2.98
P -2.62 -3.64 -3.77 -3 -1.3 -8.26 -5.15 -2.69 -2.04 -2.92 -0.75 41.01*
M
Table 7
D
The result of the sequential analysis of Forethought and Self-reflection behavior demonstrated by the LPKs.
Given/ Target G T S sE sT sS sR
TE
21.36*
EP
20
ACCEPTED MANUSCRIPT
Table 8
The result of the sequential analysis of Forethought and Self-reflection behavior demonstrated by the HPKs.
Given/ Target G T S sE sT sS sR
PT
sE 1.11 -3.65 -4.35 2.54* 12.72* -2.25 -2.53
RI
sS -5.27 -4.48 -5.35 8.46* -1.89 10.54* 7.6*
SC
sR -5.61 -4.92 -5.88 -1.14 -2.9 6.17* 21.4*
U
4.5. Overall patterns in the three phases AN
In order to have a deep understanding of differences between the LPKs and HPKs, we used the LSA
to compare their significant behavior sequences, in terms of the three phases of the SRL model (i.e.,
M
Forethought phase, Performance Phase and Self-Reflection Phase). The results (Table 9) indicated
that there were no significant differences between LPKs and HPKs in the forethought and self-
D
reflection phases but significant behavior differences were found in the performance phase. A
TE
difference among these three phases lied within the fact that students needed to set their goals and to
EP
evaluate whether they reached the goals in the forethought and self-reflection phases but they were
requested to carry out their goals in the performance phase. In other words, students needed to take
C
actions in the performance phase, instead of creating some thoughts only. This might be the reason
AC
why no significant differences existed between LPKs and HPKs in the forethought and self-
reflection phases but there were significant differences in the performance phase. Accordingly, our
finding suggested that taking actions in the performance phase was associated with students’ prior
knowledge.
21
ACCEPTED MANUSCRIPT
Table 9
Significant behaviors of LPKs and HPKs in the three phase.
Forethought phase (no significant difference between LPKs and HPKs)
● To focus on setting/modify goals, strategies and time management (G→G, T→T, S→S)
● To paid attention to evaluating time management after setting goals (G→sT)
PT
Performance Phase
Similar patterns
RI
● To follow instructions and tend to do comprehensive review (W→W, W→U, S→S, M→M)
● To view results to identify their weakness (W→R, R→R, R→C, C→C)
● To use the solutions to improve understandings (A→A, A→V, V→V)
SC
● To make notes to review and deal with learning weakness (N→N, N→P)
● To monitoring their learning status (D→U, D→S, D→D)
Different patterns
LPKs
U HPKs
AN
● To rely on notes (W/M/U→N, U→N→A) ● Less rely no notes (S→N, U→A)
● To tend to take a local approach ( U→I; U→U) ● To tend to take a global approach (M→I)
● To tend to do more to improve understandings ● No more significant behaviors
after checking learning status (D→W, D→M,
M
D→R, S→R)
D
● LPKs: Much to rely on notes (W/M/U→N, U→N→A) vs. HPKs: Less to rely no notes (S→N,
AC
U→A)
The result revealed that some differences existed between the HPKs and LPKs. The LPKs
preferred to make notes after taking the unit quiz (U→N), the weekly quiz (i.e., W→N) or the
summarized quiz (i.e., M→N). Conversely, the HPKs tended to make notes only after taking the
personal quiz. Additionally, the LPKs tended to make notes after taking the unit quiz and then
check the text solutions (U→N→A). In contrast, the HPKs did not need to take notes and they
22
ACCEPTED MANUSCRIPT
tended to check the text solutions immediately after taking the unit quiz (i.e., U→A).
These findings suggested that the LPKs relied on notes in many contexts while the HPKs
need to take notes in few contexts. The responses from the questionnaire suggested that the
LPKs were keen to make notes because they thought that making notes could help them
improve the understandings of unfamiliar topics (e.g., “I can practice unfamiliar topics more
PT
times via making notes”) and identify and overcome their weaknesses (e.g., “The function of
RI
myNote facilitated me to collect what I do not understand and remind me how to deal with
them”). On the other hand, the HPKs had sufficient prior knowledge so they did not need to rely
SC
on notes too much. In brief, making notes was essential to the LPKs while taking notes was not
●
U
LPKs: To take a local approach ( U→I; U→U) vs. HPKs: To take a global approach (M→I)
AN
The U→I for the LPKs indicated that they usually loginned the SRLSS after taking the unit
M
quiz. Such a result suggested that taking the unit quiz may be the last behavior of the LPKs
before they started to have another new session. On the other hand, the U→U suggested that
D
they repeated to take the unit quizzes. Thus, they might login the SRLSS for taking the unit
TE
quizzes only. Briefly, the LPKs were keen to take the unit quizzes when interacting with the
SRLSS. The unit quizzes focused on specific aspects so this finding suggested that the LPKs
EP
Unlike the LPKs, the M→I for the HPKs indicated that taking the summarized quiz was the
AC
last behavior of the HPKs during the process of interacting with the SRLSS. The summarized
quiz covered a variety of aspects so this finding implied that the HPKs had a tendency to take a
global approach. Such a finding was also reflected in their responses to the questionnaire. More
specifically, they attempted to take a quiz that covered more questions to confirm their
understandings. Typical responses include “It would be better if the quiz covered more difficult
questions”, “I would like to have more extensive and complex questions”. This may be able to
23
ACCEPTED MANUSCRIPT
explain why the HPKs quitted the SRLSS after taking the summarized quizzes. The
aforementioned findings echoed those of Chen, Fan and Macredie (2006), which indicated that
a global approach was appreciated by the HPKs while a local approach was favored by the
LPKs.
● LPKs: To tend to check learning status with the dashboard and quiz records ( D→W, D→M,
PT
D→R, S→R) vs. HPKs: no more significant behaviors
RI
Unlike the HPKs, the LPKs showed the patterns of the D→W and D→M, which indicated that
they usually took the weekly quiz or summarized quiz after checking the dashboard. In other
SC
words, the dashboard might motivate the LPKs to do more practices with the weekly quiz and
summarized quiz. This finding echoed that of Smits, Boon, Sluijsmans, & Van Gog (2008),
U
which addressed that the LPKs could be motivated by reminding their learning status.
AN
Therefore, they were keen to take various quizzes after checking dashboard. As addressed by
M
the LPKs “I can see my weakness via checking the dashboard. Accordingly, I can do more
practice for a single or multiple topics”, “The dashboard alerts me my weaknesses, which are
D
However, using dashboard only was not enough for the LPKs because they could get an
overview only from dashboard. Thus, they also tended to view the quiz records, which provided
EP
them with detailed situations after viewing the dashboard (D→R). Moreover, the quiz records
C
were also helpful to them after taking taking the personal quiz (S→R). This might be due to the
AC
fact that the LPKs thought that the personal quiz was a useful tool for checking their learning
status. For example, a LPK indicated that “I can check if there were any unfamiliar topics again
via taking the personal quiz and checking the answer and solution to help me understand the
concept”.
The aforementioned findings suggested that the dashboard and quiz records were useful for
LPKs because these two tools could help them identify their learning status and make
24
ACCEPTED MANUSCRIPT
improvement. However, such behavior sequences were not demonstrated by the HPKs. In other
words, the LPKs developed a significant cycle on self-monitoring and self-controlling during
the learning process. This result might also be able to explain why the learning performance of
PT
5. Discussions
RI
The results of this study revealed that prior knowledge had great effects on students’ learning
performance in a SRL context. Accordingly, we analyzed their learning behaviors during the
SC
learning process based on Zimmerman’s SRL model. Our results showed that no significant behavior
differences between the LPKs and the HPKs in the forethought phase and the self-reflection phase.
U
Nevertheless, the LPKs and HPKs demonstrated some different learning behaviors in the
AN
performance phase. That is to say, prior knowledge might greatly affect how student developed and
M
performed their learning strategies in the SRL context. The findings of this study are summarized in
25
ACCEPTED MANUSCRIPT
Fig. 8. The summary of the findings of this study
5.1. The SRL can minimize differences between the LPKs and HPKs
The result of this study revealed that the SRLSS was particularly helpful for the LPKs in improving
their learning performance. This might be because that the SRLSS provided students with various
types of support and feedback, which could help the LPKs control their learning process (Nicol, &
PT
Macfarlane‐Dick, 2006; Zamary, Rawson, & Dunlosky, 2016). Thus, the learning performance of the
RI
LPKs was considerably improved whereas that of the HPKs was not greatly changed. Such findings
echoed those found in Amadieu, Van Gog, Paas, Tricot, & Mariné (2009) and Wang, Han, Zhan, Xu,
SC
Liu, & Ren (2015).
U
5.2. The LPKs and the HPKs behave similarly in the forethought and self-reflection phases
AN
The results from the LSA indicated that the LPKs and HPKs behaved similarly in the forethought
phase and the self-reflection phase. The former was concerned with goals settings and strategies
M
planning while the latter was related to the evaluation and adjustment of strategies. In other words,
D
prior knowledge affected neither goals settings and strategies planning nor evaluation and the
TE
adjustment of strategies. This might be due to the fact that setting goals relied on self-motivation
beliefs and the awareness of current status while planning the strategies depended on the sense of
EP
self-efficacy. Such results are consistent with those of Moos & Azevedo (2008) and Taub, Azevedo,
Bouchet, & Khosravifar (2014), which showed that prior knowledge did not significantly affect
C
On the other hand, the evaluation and adjustment of strategies relied on students’ self-
reflection skills, rather than prior knowledge that they possessed. In other words, prior knowledge
has nothing to do with the evaluation and the adjustment of strategies (Bergee, 1993). Thus, the
26
ACCEPTED MANUSCRIPT
Based on the results from LSA, the LPKs and HPKs demonstrated some similarities and differences
in the performance phase. Regarding the similarities, both of them paid attention on monitoring their
learning status. This might be because understanding their own learning status could help them
PT
identify their weaknesses so that they could know where needed to be improved.
RI
Regarding the differences, the LPKs used multiple mechanisms to deal with their weaknesses
while the HPKs mainly focused on viewing the text solutions only. In other words, the LPKs and
SC
HPKs might develop different strategies to make improvement. In particular, the LPKs demonstrated
many learning behaviors after checking their learning status. These findings were coherent with
U
those of Kitsantas (2013), which suggested that prior knowledge might affect student’s learning
AN
behaviors (Shen, & Chen, 2006; Cifarelli, Goodson-Espy, & Chae, 2010). This might be the reason
M
why the LPKs and the HPKs demonstrated different learning behaviors in this phase.
D
6. Conclusions
TE
This study investigates “whether the proposed SRL environment is useful for students to improve
learning performance”, “how LPKs and the HPKs behaved differently in the SRL environment?” and
EP
“why LPKs and the HPKs behaved differently”. To answer these three research questions, the
C
learning performance, sequential learning behaviors and qualitative data from the questionnaires
AC
were considered. Regarding learning performance, the results from the delayed test showed that no
significant difference exist between the LPKs and HPKs. In other words, the LPKs’ performance
was improved after a long-term learning process. Accordingly, these findings implied that the SRL
was helpful to remove the gap of learning performance caused by prior knowledge.
Regarding learning behaviors, each event was coded and classified based on the three phases
of the SRL model, i.e., Forethought phase, Performance Phase and Self-Reflection Phase. The
27
ACCEPTED MANUSCRIPT
results of this study suggested that no significant difference between the HPKs and LPKs existed in
the forethought phase and self-reflection phase. However, prior knowledge had great effects on
students’ behaviors in the performance phase, where the LPKs and HPKs demonstrated different
learning behaviors. In summary, the LPKs, who did not possess sufficient prior knowledge, tended to
take a local approach with the unit quiz, while the HPKs, who had enough prior knowledge, had a
PT
tendency to take a global approach with the summarized quiz. In addition, the lack of prior
RI
knowledge might also make the LPKs more rely on the note while the HPKs did not demonstrate
such a pattern. Furthermore, the LPKs tended to check their learning status with the dashboard and
SC
quiz records and subsequently attempted to make some improvement with various quizzes. This
might be due to the fact that the LPKs lacked self-monitoring and self-controlling strategies so they
U
needed to seek external support provided by the dashboard and quiz records and then did some
AN
practices with several quizzes. Conversely, the HPKs possessed such strategies so they did not need
M
to obtain additional support and did some practice. Such findings also echoed results from previous
studies, which indicated that external support might be beneficial to the LPKs (e.g., Winters, Greene,
D
& Costich, 2008; Azevedo, Moos, Greene, Winters, & Cromley, 2008).
TE
The contribution of this study includes three aspects: theory, methodology, and applications.
In terms of theory, this study deepens the understandings of the impacts of the SRL approaches on
EP
student learning by providing empirical evidence. The findings of this study indicated that providing
C
students with the SRL support was helpful for LPKs to develop an adaptive learning strategy, with
AC
which they tended to master what they had learnt so their learning performance could be improved.
However, the sample size was not big in this study. Further work needs to be undertaken with a
With regard to methodology, this study analyzed the students’ learning behavior with the
LSA approach and qualitative data. Such approaches are useful to get a further understanding of
28
ACCEPTED MANUSCRIPT
conduct further work to analyze students’ learning behaviors with other approaches, e.g., a data
mining approach. It would be interesting to see whether similar results could be obtained by using
such an approach. As far as the application is concerned, this study illustrates how to implement a
SRL that can accommodate students’ individual differences. However, this study focused on prior
knowledge only and other individual differences (e.g., genders, cognitive styles) might also have
PT
great effects. Thus, such individual differences should be addressed in our future works. The findings
RI
from such future works can be integrated with those from this present work. By doing so, we can
develop personalized SRL that can satisfy the needs and preferences of each individual.
SC
References
U
Amadieu, F., Van Gog, T., Paas, F., Tricot, A., & Mariné, C. (2009). Effects of prior knowledge and
AN
concept-map structure on disorientation, cognitive load, and learning. Learning and Instruction,
M
19(5), 376-386.
satisfaction with online training. Journal of computer assisted learning, 24(3), 260-270.
TE
Azevedo, R., Moos, D. C., Greene, J. A., Winters, F. I., & Cromley, J. G. (2008). Why is externally-
facilitated regulated learning more effective than self-regulated learning with hypermedia?.
EP
Bellhäuser, H., Lösch, T., Winter, C., & Schmitz, B. (2016). Applying a web-based training to foster
Bergee, M. J. (1993). A comparison of faculty, peer, and self-evaluation of applied brass jury
29
ACCEPTED MANUSCRIPT
Bernacki, M. L., Byrnes, J. P., & Cromley, J. G. (2012). The effects of achievement goals and self-
Bernacki, M. L., Byrnes, J. P., & Cromley, J. G. (2012). The effects of achievement goals and self-
PT
environments. Contemporary Educational Psychology, 37(2), 148-161.
RI
Biswas, G., Jeong, H., Kinnebrew, J. S., Sulcer, B., & ROSCOE, R. (2010). Measuring self-
regulated learning skills through social interactions in a teachable agent environment. Research
SC
and Practice in Technology Enhanced Learning, 5(02), 123-152.
Broadbent, J. (2017). Comparing online and blended learner's self-regulated learning strategies and
U
academic performance. The Internet and Higher Education, 33, 24-32.
AN
Cifarelli, V., Goodson-Espy, T., & Chae, J. L. (2010). Associations of students’ beliefs with self-
M
regulated problem solving in college algebra. Journal of Advanced Academics, 21(2), 204-232.
Chen, S. Y., Fan, J. P., & Macredie, R. D. (2006). Navigation in hypermedia learning systems:
D
Cleary, T. J., & Zimmerman, B. J. (2001). Self-regulation differences during athletic practice by
experts, non-experts, and novices. Journal of Applied Sport Psychology, 13(2), 185-206.
EP
program to enhance self‐regulated and self‐motivated cycles of student learning. Psychology in the
AC
Dochy, F., Segers, M., & Buehl, M. M. (1999). The relation between assessment practices and
outcomes of studies: The case of research on prior knowledge. Review of educational research,
69(2), 145-186.
30
ACCEPTED MANUSCRIPT
Dunlap, J. C. (2005). Problem-based learning and self-efficacy: How a capstone course prepares
students for a profession. Educational Technology Research and Development, 53(1), 65-83.
Ekholm, E., Zumbrunn, S., & Conklin, S. (2015). The relation of college student self-efficacy
toward writing and writing self-regulation aptitude: writing feedback perceptions as a mediating
PT
Fisher, S. L., & Ford, J. K. (1998). Differential effects of learner effort and goal orientation on two
RI
learning outcomes. Personnel Psychology, 51(2), 397-420.
Greene, J. A., & Azevedo, R. (2009). A macro-level analysis of SRL processes and their relations to
SC
the acquisition of a sophisticated mental model of a complex system. Contemporary Educational
U
Greene, J. A., Moos, D. C., Azevedo, R., & Winters, F. I. (2008). Exploring differences between
AN
gifted and grade-level students’ use of self-regulatory learning processes with hypermedia.
M
Goetz, T., Preckel, F., Pekrun, R., & Hall, N. C. (2007). Emotional experiences during test taking:
D
Does cognitive ability make a difference?. Learning and Individual Differences, 17(1), 3-16.
TE
Guskey, T. (1980). Mastery learning: Applying the theory. Theory into Practice, 19(2), 104-111.
Kapa, E. (2001). A metacognitive support during the process of problem solving in a computerized
EP
Kramarski, B., & Gutman, M. (2006). How can self‐regulated learning be supported in mathematical
AC
31
ACCEPTED MANUSCRIPT
Lehmann, T., Hähnlein, I., & Ifenthaler, D. (2014). Cognitive, metacognitive and motivational
313-323.
Liu, H. C., Andre, T., & Greenbowe, T. (2008). The impact of learner’s prior knowledge on their use
PT
Technology, 17(5), 466-482.
RI
Mitchell, T. J., Chen, S. Y., & Macredie, R. D. (2005). Hypermedia learning and prior knowledge:
domain expertise vs. system expertise. Journal of Computer Assisted Learning, 21(1), 53-64.
SC
Moos, D. C., & Azevedo, R. (2008). Self-regulated learning with hypermedia: The role of prior
U
Nicol, D., & Macfarlane-Dick, D. (2006). Formative assessment and self-regulated learning: a
AN
model and seven principles of good feedback practice. Studies in Higher Education, 31(2), 199-
M
218.
Pintrich, & M. Zeidner (Eds.), Handbook of self-regulation (pp. 451–502). San Diego, CA:
TE
Academic.
Roll, I., Baker, R. S. D., Aleven, V., & Koedinger, K. R. (2014). On the benefits of seeking (and
EP
avoiding) help in online problem-solving environments. Journal of the Learning Sciences, 23(4),
C
537-560.
AC
Schunk, D.H. (2005) Self-regulated learning: the educational legacy of Paul R. Pintrich. Educational
Shen, B., & Chen, A. (2006). Examining the interrelations among knowledge, interests, and learning
32
ACCEPTED MANUSCRIPT
Smits, M. H., Boon, J., Sluijsmans, D. M., & Van Gog, T. (2008). Content and timing of feedback in
Song, H. S., Kalet, A. L., & Plass, J. L. (2016). Interplay of prior knowledge, self‐regulation and
PT
Learning, 32(1), 31-50.
RI
Spruce, R., & Bol, L. (2015). Teacher beliefs, knowledge, and practice of self-regulated
SC
Sung, Y. T., Liao, C. N., Chang, T. H., Chen, C. L., & Chang, K. E. (2016). The effect of online
summary assessment and feedback system on the summary writing on 6th graders: The LSA-
U
based technique. Computers & Education, 95, 1-18.
AN
Taub, M., Azevedo, R., Bouchet, F., & Khosravifar, B. (2014). Can the use of cognitive and
M
Trevors, G., Duffy, M., & Azevedo, R. (2014). Note-taking within MetaTutor: interactions between
TE
an intelligent tutoring system and prior knowledge on note-taking and learning. Educational
Tsai, C. C., Lin, S. S., & Yuan, S. M. (2001). Students' use of web‐based concept map testing and
C
Wang, D., Han, H., Zhan, Z., Xu, J., Liu, Q., & Ren, G. (2015). A problem solving oriented
intelligent tutoring system to improve students' acquisition of basic computer skills. Computers &
Wang, T. H. (2011). Developing Web-based assessment strategies for facilitating junior high school
33
ACCEPTED MANUSCRIPT
Winne, P. H. (1995). Inherent details in self-regulated learning. Educational psychologist, 30(4),
173-187.
Winters, F. I., Greene, J. A., & Costich, C. M. (2008). Self-regulation of learning within computer-
based learning environments: A critical analysis. Educational Psychology Review, 20(4), 429-444.
Zamary, A., Rawson, K. A., & Dunlosky, J. (2016). How accurately can students evaluate the
PT
quality of self-generated examples of declarative concepts? Not well, and feedback does not
RI
help. Learning and Instruction, 46, 12-20.
SC
Relating grade, sex, and giftedness to self-efficacy and strategy use. Journal of educational
U
Zimmerman, B. J. (2000). Attaining self-regulation: A social cognitive perspective. In M. Boekaerts,
AN
P. R. Pintrich, & M. Zeidner (Eds.), Handbook of self-regulation (pp. 13–39). San Diego, CA:
M
Academic Press.
41(2), 64–70.
TE
45(1), 166-183.
C
Zimmerman, B. J., & Kitsantas, A. (2007). Reliability and validity of Self-efficacy for Learning
AC
34
ACCEPTED MANUSCRIPT
Highlights
A self-regulated learning support system (SRLSS) was proposed.
The SRLSS could remove the gap between the High Prior Knowledge students
and Low Prior Knowledge students.
Prior knowledge affects students’ behavior in the performance phrase.
PT
RI
U SC
AN
M
D
TE
C EP
AC