Motivated Strategies For Learning Questionnaire MSLQ

15206807, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/pits.22845 by University Of Macau Wu Yee Sun Library (E2), Wiley Online Library on [30/12/2022].
See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Received: 27 April 2022 | Revised: 26 October 2022 | Accepted: 1 December 2022
DOI: 10.1002/pits.22845
RESEARCH ARTICLE
Motivated Strategies for Learning Questionnaire

(MSLQ): Adaptation, validation, and
development of a short form in the Chinese
context for mathematics
Faming Wang1 | Chunlian Jiang2 | Ronnel B. King3 |

Shing On Leung2
1
Centre for the Enhancement of Teaching
and Learning, The University of Hong Kong, Abstract
Hong Kong SAR, China
The Motivated Strategies for Learning Questionnaire (MSLQ)
2
Faculty of Education, University of Macau,
Macau SAR, China
is one of the most widely used instruments to measure
3
Department of Curriculum and Instruction, students' motivation and self‐regulated learning. However, the
Faculty of Education, The Chinese University MSLQ was developed and has been predominantly used in
of Hong Kong, Hong Kong SAR, China
the Western context, is a domain‐general measure, and is
Correspondence quite lengthy. Hence, adapting the MSLQ to the Chinese
Faming Wang, Centre for the Enhancement
of Teaching and Learning, The University of
educational context, validating its application in specific
Hong Kong, Hong Kong SAR, China. subjects, and developing a short form would be an optimal
Email: [email protected]
way to improve its accessibility. This study aimed to
investigate the psychometric properties of the Chinese version
of the MSLQ in mathematics learning (i.e., MSLQ‐C) and
develop a short form (i.e., MSLQ‐CS) using set exploratory
structural equation modeling. The sample consisted of 563
senior secondary students in China. Results demonstrated that
both MSLQ‐C and MSLQ‐CS showed acceptable construct
validity, reliability, and concurrent validity. Furthermore,
structural relationships and interrelationships among the
subscales and their relationships with mathematics achieve-
ment were highly similar for MSLQ‐C and MSLQ‐CS.
Theoretical and practical implications are discussed.
KEYWORDS
Chinese version, mathematics learning, Motivated Strategies for
Learning Questionnaire (MSLQ), set exploratory structural
equation modeling (Set‐ESEM), short form
Psychology in the Schools. 2022;1–23. wileyonlinelibrary.com/journal/pits © 2022 Wiley Periodicals LLC. | 1

15206807, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/pits.22845 by University Of Macau Wu Yee Sun Library (E2), Wiley Online Library on [30/12/2022]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
2 | WANG ET AL.
1 | INTRODUCTION
In the 21st century, knowledge and mastery of mathematics are becoming increasingly critical (Gravemeijer et al., 2017).
In the coming decade, occupations related to science, technology, engineering, and math (STEM) will see a rapid increase
and grow at twice the rate compared to non‐STEM occupations (Zilberman & Ice, 2021). However, mathematics is often
perceived as a difficult subject, with many students experiencing motivational problems and using suboptimal learning
strategies. For example, international studies have shown that 41% of eighth‐grade students do not like learning
mathematics, and 44% of students are not confident in learning mathematics (Mullis et al., 2020).
Existing research has consistently highlighted the critical role of self‐regulated learning in facilitating
mathematics achievement (e.g., Perels et al., 2009; Robson et al., 2020). Measuring students' learning motivation
and strategies with a psychometrically sound tool should be an important leverage point for improving mathematics
achievement. The Motivated Strategies for Learning Questionnaire (MSLQ), developed by Pintrich et al. (1991),
might be useful in this regard. The MSLQ has been extensively used to understand students' self‐regulated learning,
such as exploring individual differences in motivation and evaluating the effectiveness of interventions targeting
learning strategies (Duncan & McKeachie, 2005).
However, there are two gaps in terms of the questionnaire itself. First, the psychometric properties of the
MSLQ (e.g., factor structure, reliability, and construct validity) vary across countries, subjects, and grade levels,
demonstrating the necessity of adapting and validating the MSLQ with diverse populations (Pintrich et al., 2000). In
a meta‐analytic study, Credé and Phillips (2011) reviewed 59 studies that adapted the MSLQ in several countries
(e.g., the United States and Canada) across a wide variety of subjects (e.g., mathematics and English). They found
that the MSLQ is a reliable instrument; however, the psychometric properties of items and variables need further
modification to obtain better criterion validity and factor structure. To our knowledge, little research has been
conducted to explore the psychometric properties of the MSLQ in the Chinese context, especially for mathematics
learning. Such research is of great significance in terms of the critical role of mathematics in the 21st century.
Meanwhile, Chinese secondary school students face fierce competition in the Chinese National Higher Education
Entrance Examination, where mathematics is a major subject and occupies a relatively large portion of the total
score. As Kennedy and Lee (2007) argued, examination systems in China follow the traditional culture of
“assessment for selection”. The test score would determine the ranking of the university that students can enter
and subsquently influence their career development. Hence, it is important to establish a valid and reliable
instrument to measure students' motivation and learning strategies in mathematics. Second, the MSLQ was
designed based on Pintrich's integrated self‐regulated learning framework and aimed to measure various
motivational, cognitive, and behavioral strategies (Pintrich et al., 1993). However, the broad scope entails the use of
a large number of items in the MSLQ (i.e., 81 items), which may result in survey fatigue and weakened quality and
effectiveness of measurement. More importantly, the large space accounted for by the lengthy MSLQ items makes
it unrealistic to simultaneously measure self‐regulated variables and other constructs, preventing researchers from
exploring complex research questions that might interest them (Ziegler et al., 2014).
Given the above gaps, the current study has two purposes. The first is to adapt and validate the Chinese version
of the MSLQ for the mathematics domain (i.e., MSLQ‐C). The second is to develop and validate a short form of the
MSLQ (i.e., MSLQ‐CS). Furthermore, considering that the MSLQ includes multiple related variables of the same
domain, set exploratory structural equation modeling (Set‐ESEM) is carried out to explore the factor structure of the
MSLQ‐C and the MSLQ‐CS (Marsh et al., 2020).
1.1 | Self‐regulated learning and mathematics learning
Self‐regulated learning is an umbrella concept that involves a set of conscious and effortful awareness, achievable
task‐specific targets and cognitive goals, coordination of various cognitive strategies, and regulation of behaviors
WANG ET AL. | 3
and contextual factors (Cleary & Zimmerman, 2004; Duncan & McKeachie, 2005). Pintrich (2000, p. 453) described
self‐regulated learning as “an active, constructive process whereby learners set goals for their learning and then
attempt to monitor, regulate and control their cognition, intentions, and behavior, guided and constrained by their
goals and the contextual features of the environment.” Pintrich's self‐regulated learning model posits that students'
ability to regulate cognition, motivation/affect, behavior, and context can be learned, controlled, and regulated
(Pintrich & Zusho, 2002; Pintrich, 2004; Schunk, 2005). Accordingly, self‐regulated learners are the ones who can
successfully activate and maintain their motivational beliefs, adopt adaptive learning cognitive and metacognitive
strategies, and control overt behaviors (Pintrich, 2004).
In the domain of mathematics, the critical role of self‐regulated learning strategies in mathematics learning has
been well‐documented. For example, Cleary et al. (2021) classified students into five profiles based on their self‐
regulated learning variables (e.g., students' perceived use of regulatory strategies and self‐efficacy beliefs to engage
in self‐regulated learning) and found that students in the adaptive self‐regulated learning profile tended to have a
higher level of classroom engagement and mathematics achievement than their peers. In another study, DiGiacomo
and Chen (2016) indicated that the self‐regulated learning intervention incorporating monitoring and self‐regulation
strategies into the math curriculum could facilitate mathematics achievement.
1.2 | MSLQ
The MSLQ is a self‐reported questionnaire developed by Pintrich et al. (1991, 1993) to measure students' self‐
regulation, which comprises 81 Likert‐scale items with five sets. The first three sets were designed to measure
motivational beliefs, including expectancy, value, and affect. Expectancy refers to students' beliefs in accomplishing
the given tasks, represented by self‐efficacy (i.e., students' judgment on their ability to perform a task) and control
beliefs for learning (i.e., students' beliefs about whether their effort can result in positive outcomes). Value focuses
on why students engage in academic tasks, including intrinsic goal orientation (i.e., doing tasks for reasons of
learning and mastery), extrinsic goal orientation (i.e., doing tasks for reasons of rewards, grades, and competition
with others), and task value beliefs (i.e., beliefs about the interest, importance, and utility of tasks). Affect is
operationalized as test anxiety (i.e., students' feelings of panic and fear regarding the test) and measured by
students' worry about taking the exam.
The fourth set was used to measure students' cognitive and metacognitive strategies. Cognitive strategies
include a series of basic to complex strategies that students use to process information in learning, including
rehearsal (e.g., reciting or naming items from a list to be learned), elaboration (e.g., paraphrasing and summarizing),
organization (e.g., selecting appropriate information), and critical thinking (e.g., applying previous knowledge to new
situations). Metacognitive strategies are operationalized as students' use of strategies that help them control and
regulate their cognition.
The last set was resource management, which refers to students' regulatory strategies for controlling other
resources besides their cognition, including managing time and study environment (e.g., managing study time),
effort regulation (e.g., persisting in the face of difficulties), peer learning (e.g., learning in a group with peers), and
help‐seeking (e.g., seeking help from peers when needed).
The MSLQ is one of the most promising tools to measure students' self‐regulated learning for several reasons.
First, the MSLQ is developed based on Pintrich's self‐regulated learning framework and is an integrative instrument
that measures various facets of self‐regulated learning, including cognitive, motivational, and behavioral–contextual
elements (Panadero, 2017). Second, the MSLQ can be flexibly applied to various research questions on self‐
regulated learning, given that all variables can be used either jointly or separately (Duncan & McKeachie, 2005).
Third, many studies have indicated that MSLQ is an efficient, practical, and ecologically valid measure of students'
self‐regulated learning across cultures, countries, samples, and subjects (e.g., Meijs et al., 2019; Rao & Sachs, 1999;
Wong et al., 2013). Fourth, the MSLQ is designed at the course level with appropriate grain size, avoiding either too
4 | WANG ET AL.
much generality (e.g., all learning situations) or too much specificity (e.g., specific situations within one course)
(Duncan & McKeachie, 2005).
1.3 | Psychometric properties of MSLQ
Pintrich and Zusho (2002) demonstrated that self‐regulated learning was context‐specific and influenced by the
interaction of person, behavior, and environment. A large body of research indicates that students' self‐regulation is
subject to different cultures (King & McInerney, 2014, 2016; Mclnerney & King, 2018), domains (Wolters &
Pintrich, 1998), and grade levels (Dent & Koenka, 2016).
Considering the influence of contexts, researchers have validated the psychometric properties of the MSLQ in
many countries (e.g., Australia, Germany, and Korea), subjects (e.g., mathematics, language, and science), and
samples at different grade levels (e.g., middle school students, high school students, and graduate students) (see
Bonanomi et al., 2018; Duncan & McKeachie, 2005 for reviews). It was reported that the psychometric properties
of the MSLQ varied in different contexts (e.g., Hilpert et al., 2013; Jackson, 2018; Pintrich & de Groot, 1990;
Pintrich et al., 1991). For example, Hilpert et al. (2013) administered the MSLQ to postsecondary students enrolled
in introductory geoscience courses. They retained six MSLQ variables (i.e., self‐efficacy, control of learning, intrinsic
goals, task value, metacognitive regulation, and effort regulation) and fit a three‐factor structure model (i.e.,
expectancy, value, and self‐regulation). Meijs et al. (2019) investigated the underlying structure of the learning
strategy section for university students who majored in distance education. A five‐factor structure was found,
including management of time and effort, complex cognitive strategy use, simple cognitive strategy use, contacts
with others, and academic thinking.
However, relatively fewer studies have investigated the psychometric properties of the MSLQ in the Chinese
context (e.g., Lee et al., 2010; Sachs et al., 2001; Tong et al., 2019; Wang et al., 2021; Xu, 2020). The targeted
samples of these studies were limited to junior secondary school students (e.g., Lee et al., 2010; Xu, 2020), a
mixture of primary and junior secondary school students (e.g., Sachs et al., 2001), or college students (e.g., Tong
et al., 2019). Furthermore, these studies focused on the application of the MSLQ in the general domain (e.g., Tong
et al., 2019; Zhou & Wang, 2021). Although the study conducted by Xu (2020) is rooted in mathematics, only three
cognitive variables (e.g., organization, elaboration, and critical thinking) were adapted and validated. To conclude,
studies that specifically explored the psychometric properties of the MSLQ for Chinese senior secondary school
students and focused on mathematics are relatively limited.
1.4 | The need for a short form of MSLQ
There are at least three advantages to applying the short form in research. First, a short questionnaire allows
researchers to simultaneously measure more constructs (Ziegler et al., 2014). Second, a short questionnaire is easier
to administer given that it saves response time, maximizes the utility of questionnaire space, and has fewer logistical
issues. Third, a short questionnaire may mitigate the problems caused by participants' feelings of fatigue and
boredom, reducing missing data produced by careless answers (Credé et al., 2012).
Researchers have attempted to develop a short form of MSLQ by removing some of the variables (e.g., Pintrich
& de Groot, 1990; Rao & Sachs, 1999). For example, Pintrich and de Groot (1990) developed a short form for junior
high school students by keeping five variables (i.e., self‐regulation, cognitive strategy use, self‐efficacy, intrinsic
value, and test anxiety). However, these studies created a short form by removing some variables from the original
MSLQ rather than selecting a subset of items within each variable, making it difficult to compare whether the short
form is better than its long form. Second, the short form with selected variables cannot measure the whole
complexity of motivation and learning strategies proposed in Pintrich's self‐regulation framework. Hence, this study
WANG ET AL. | 5
aims to develop a short form of the MSLQ by reducing items within each variable and keeping the full range number
of variables included in the original MSLQ.
1.5 | Set‐exploratory structural equation modeling (Set‐ESEM)
In the existing studies, exploratory factor analysis (EFA), confirmatory factor analysis (CFA), and ESEM have been
widely used in examining the factor structure of the MSLQ (e.g., Hilpert et al., 2013; Thomas & Cassady, 2019; Tong
et al., 2020; Xu, 2020). However, these studies might be beset by some limitations. EFA is utterly data‐driven and is
not aimed at fitting any pre‐designed models. For CFA, the independent clusters model assumption is too restrictive
for multidimensional constructs, which is often not available in practice with the poor model fit (Morin et al., 2013).
Meanwhile, ESEM lacks parsimony (especially applied in complex models with many constructs based on small‐to‐
moderate sample sizes, such as MSLQ) and might confound constructs that need to be separated theoretically
(Marsh et al., 2020).
Set‐ESEM might be able to address these problems. In Set‐ESEM, items are first divided into different sets, with
different factors within a set. Then, item loadings across different sets are restricted to be zero, while item loadings
across different factors within the same set are constrained to be close to zero. Finally, item loadings within the
same factor are freely estimated (Dicke et al., 2018). Hence, set‐ESEM is a compromise between CFA and ESEM
regarding goodness‐of‐fit, parsimony (regarding the number of estimated parameters), rigor (based on model
constraints), and factor structure (i.e., the empirical mapping of items to factors that corresponds to the a priori
theoretical model) (Marsh et al., 2020). To date, Set‐ESEM has been popular in many research fields. For example,
Dicke et al. (2018) applied Set‐ESEM to validate the Copenhagen Psychosocial Questionnaire. Yukhymenko‐
Lescroart and Gilbert (2021) adopted Set‐ESEM to develop and validate the Coaching Athlete Purpose Scale.
Set‐ESEM is especially appropriate for investigating the factor structure of MSLQ. The whole MSLQ
questionnaire can be grouped into five sets for measuring different facets of self‐regulated learning, including value,
expectancy, affect, cognitive and metacognitive strategies, and resource management strategies. Identifying these
five sets in exploring the factor structure of the MSLQ is vital. The five sets of factors are different from each other.
Hence, item loadings across sets are restricted to be zero. Meanwhile, item loadings across different factors within a
set are allowed to be non‐zero but small, given that variables within each subset are closely related.
1.6 | The Chinese context
The present study focused on Chinese senior secondary students, who are influenced by Confucian culture and
experience tremendous pressure in preparing for the Chinese National Higher Education Entrance Examination.
Generally, Chinese students tend to study in a teacher‐directed, exam‐oriented, and utilitarian learning environment
(Hwang, 2015; Zhang et al., 2004). It has been reported that Chinese students are quiet, receptive, and disinclined
to challenge authority (Pratt et al., 1999). Teachers get accustomed to adopting strategies that teach to the exam
rather than fostering deep learning, which could thwart students' self‐regulated learning (Zhao, 2020). For example,
Chinese mathematics teachers tend to implement the traditional didactic approach as the primary teaching strategy
and view the repeated practice as a key to consolidating knowledge (Wang & Cai, 2007). Hence, Chinese students
tend to use more surface learning strategies (e.g., rehearsal) than their counterparts in western countries
(Leung, 2001).
Meanwhile, motivational factors could interact with cognitive, behavioral, and environmental factors to
influence self‐regulated learning (King, 2022; Schunk, 2005). For example, existing studies have documented
that students' intrinsic and extrinsic goal orientation are closely associated with self‐regulated learning strategies
(Cai et al., 2022; Haw & King, 2022; Wolters et al., 1996). Evidence from Western cultures indicated that intrinsic
6 | WANG ET AL.
goal orientation generally facilitates the use of sophisticated strategies, while extrinsic goal orientation makes
students feel controlled and hinders self‐regulated learning (e.g., Pintrich & de Groot, 1990; Walker et al., 2006).
However, the influence of motivational factors on Chinese students' self‐regulated learning might differ. For
example, Chinese students tend to have high attainment and utility value for learning due to the importance of
education in social mobility (Li, 2003). Hence, researchers indicated that extrinsic goal orientation plays a more
critical role than intrinsic goal orientation for Chinese students (Hau & Ho, 2010; Wang et al., 2019). These findings
suggest the need to explore self‐regulated learning (i.e., motivational beliefs and learning strategies) among Chinese
senior secondary students.
1.7 | The present study
Two research goals directed this study. The first goal is to adapt MSLQ in the Chinese context for mathematics
(i.e., MSLQ‐C) and scrutinize its psychometric properties, including construct validity, reliability, and concurrent
validity. The second goal is to develop the short form (i.e., MSLQ‐CS) based on the adapted Chinese version of
MSLQ and assess its psychometric properties by conducting systematic comparisons with the MSLQ‐C.
2 | M E TH O D S
2.1 | Participants
The sample was selected using a convenience sampling method from Guangdong province, an economically
advanced province in China. First, Soper's (2022) online SEM calculator, developed based on Cohen (1988) and
Westland (2010), was used to determine the minimum sample size in the present study. The result indicated
that a sample size of 498 was adequate to detect a small to moderate effect size (0.3), assuming a 95% power
and a 5% level of statistical significance. Second, several invitation letters were sent to the principals. Third,
within each school whose principal accepted, we randomly invited two classes of Grade 11 students to
participate in our research. Consent forms were collected before data collection, and all participants were
informed of the research purposes and procedures. Finally, we obtained 563 senior secondary school students
from four schools. The average age of the participants was 17.12, with a standard deviation of 0.48. There were
259 females (46.6%).
2.2 | Procedures for translating MSLQ
Following the International Test Commission (2018) guidelines for test adaptation, items were translated into
Chinese and adapted to the subject of mathematics using a committee approach. The team consisted of an
educational measurement expert, a mathematics educationalist, and a PhD candidate majoring in educational
assessment in mathematics education. The committee members are equipped with the combined knowledge of (1)
the languages involved (i.e., English and Chinese), (2) the Chinese cultures, (3) the Chinese mathematics education,
and (4) the knowledge of educational measurement and assessment.
According to the recommendation of the Organization for Economic Co‐operation and Development
(OECD, 2020), a double‐translation and reconciliation procedure was applied. First, the educational measurement
expert and PhD candidate conducted the forward translation independently to translate the questionnaire from
English to Chinese. Second, the discrepancies were recorded and reconciled by the mathematics educationalist.
Third, the modification was conducted after the discussion with a consensus on each item.
WANG ET AL. | 7
2.3 | Measures
2.3.1 | MSLQ
The five sets with full 81‐item MSLQ were measured in the present study. First, the expectancy set comprises
control of learning beliefs (e.g., “It is my own fault if I don't learn the material in the math course.”) and self‐efficacy
for mathematics learning and performance (e.g., “I believe I will receive an excellent grade in the math class.”).
Second, the value set includes intrinsic goal orientation (e.g., “In the math class, I prefer course material that really
challenges me so I can learn new things.”), extrinsic goal orientation (e.g., “Getting a good grade in the math class is
the most satisfying thing for me right now.”), and task value (e.g., “I think I will be able to use what I learn in the math
course in other courses.”). Third, the affect set is composed of test anxiety (e.g., “When I take math tests I think if
the consequences of failing.”). Fourth, the set of cognitive and metacognitive strategies includes rehearsal (e.g., “I
make lists of important terms for the math course and memorize the lists.”), elaboration (e.g., “I try to apply ideas
from the math course readings in other class activities such as lecture and discussion.”), organization (e.g., “I make
simple charts, diagrams, or tables to help me organize math course material.”), critical thinking (e.g., “I treat the math
course material as a starting point and try to develop my own ideas about it.”), and metacognitive self‐regulation
(e.g., “If I get confused taking notes in math class, I make sure I sort it out afterwards.”). Fifth, the set of resource
management strategies comprises time and study environment management (e.g., “I make good use of my study
time for the math course.”), effort regulation (e.g., “I work hard to do well in the math class even if I don't like what
we are doing.”), peer learning (e.g., “When studying for the math course, I often try to explain the material to a
classmate or a friend.”), and help seeking (e.g., “I try to identify students in the math class whom I can ask for help if
necessary.”).
The original MSLQ was designed to measure undergraduates' learning motivation and strategies on a “Learning
to Learn” course at the University of Michigan. In the original MSLQ, all items were rated on a 7‐point Likert scale,
from 1 (Not at all true of me) to 7 (Very true of me). The reliability of the variables ranged from 0.52 to 0.93. The
confirmatory analyses showed that MSLQ showed sound structure and good factor loadings. However, the
goodness‐of‐fit was not stellar in the original MSLQ as the sample was from various courses and subject domains
(Pintrich et al., 1991).
In this study, items were adapted and answered on an 11‐point Likert scale from 0 (Not at all true of me) to 10
(Very true of me), given that 11‐point scale is closer to the interval level of scaling and normality (Leung, 2011).
Previous studies that adapted and validated the MSLQ for Chinese students have found that it has acceptable
reliability and validity in the Chinese context (e.g., Lee et al., 2010; Sachs et al., 2001; Tong et al., 2019; Zhou &
Wang, 2021). For example, Tong et al. (2019) adapted the MSLQ for Chinese undergraduate students and found the
overall reliability coefficient for motivation and learning strategies was 0.80 and 0.85, respectively. The fit indices of
factor structure were also acceptable. Meanwhile, these studies suggested that the psychometric properties of
MSLQ differed in diverse contexts regarding the subjects and students' grade levels.
2.3.2 | Mathematics achievement
We developed a practice test based on the Chinese National Higher Education Entrance Examination test to
measure students' basic mathematics knowledge and skills based on the curriculum standard. Two steps were
conducted to ensure the content validity of the test. First, we examined the test specifications to scrutinize
the nature of the examination, cognitive framework, topics to be covered, and the cognitive requirement for
each topic. The cognitive frameworks have three levels of knowing, understanding, and mastering, and the
skills to be evaluated, including spatial imaginary skills, abstract thinking skills, reasoning/deductive skills,
computational skills, data handling skills, and creative and application skills. Second, we determined the
8 | WANG ET AL.
content areas through an evaluation of each item in the test. Finally, we developed 12 multiple‐choice items,
four short‐answer questions, and four open‐response items (see Table S1 for the contents covered in the test).
The scores were converted to a scale of 0–100. This test demonstrated good internal consistency with
Cronbach's α coefficients of .93.
2.4 | Data analyses
The percentage of missing data was low, ranging from 0% to 1.4%. Full Information Maximum Likelihood estimation
was used to handle missing data. This method led to unbiased parameter estimates and is superior to other methods
such as listwise deletion, pairwise deletion, and mean substitution (Enders, 2010).
2.4.1 | Phase 1: Adapting and validating the long form: MSLQ‐C
2.4.1.1 | Initial examination of item performance

The mean, standard deviation (SD), skewness, kurtosis, and item‐total correlations were computed to examine the
item performance. Univariate normality was assessed by Kline's (2005) criteria, following the recommended values
that skewness and kurtosis should be under |3| and |10|, respectively.
2.4.1.2 | Construct validity

Set‐ESEM with target rotation was used to investigate the construct validity of the MSLQ‐C with Mplus 8.0
(Muthén & Muthén, 2017). Based on the factor structure proposed by Pintrich et al. (1991), Set‐ESEM analyses with
five‐set factors were conducted. Figure 1 shows the conceptual model.
The classical goodness‐of‐fit indices were used to assess the model fit. The value of the comparative fit index
(CFI) and Tucker–Lewis index (TLI) greater than 0.90 and root mean square error of approximation (RMSEA) less
than 0.08 were interpreted as acceptable fit (Marsh et al., 2004).
F I G U R E 1 Factor structure of MSLQ based on Set‐ESEM method. Note: The correlations between latent
variables were not shown in the figure for parsimony. The solid lines indicate that the items are mainly loaded on
the factor, while the dotted lines indicate the items cross‐loaded on factors. CLF, control of learning beliefs; CT,
critical thinking; EGO, extrinsic goal orientation; ELA, elaboration; HS, help seeking; IGO, intrinsic goal orientation;
MSR, metacognitive self‐regulation; ORG, organization; PL, peer learning; RE, rehearsal; SE, self‐efficacy for
learning and performance; TA, test anxiety; TE, time and study environment management; TV, task value.
WANG ET AL. | 9
2.4.1.3 | Reliability
We used the coefficient omega (ω) to assess reliability because Cronbach's alpha reliabilities have been found to
underestimate scale reliability. Cronbach's alpha was built on an essentially tau‐equivalent model assumption,
where the factor loadings were required to be equivalent for all items (Dunn et al., 2014; Lucke, 2005). However,
this assumption was difficult to achieve in the Set‐ESEM model. In comparison, ω allowed the factor loadings to
vary across items (Graham, 2006). ω can be understood as the proportion of the variance explained by a latent
construct relative to the observed‐score variance (Dunn et al., 2014). The cutoff values for ω were the same as
Cronbach's alpha. Values of ω greater than 0.7 were considered acceptable, while those above 0.8 were regarded as
high (Kline, 2000). However, all cutoff values about reliabilities are arbitrary, and “the nature of the decision being
made on the basis of a test should be the guide for the acceptable level of reliability” (Cho & Kim, 2015; p.218). The
values of ω ranging from 0.65 to 0.7 were also considered acceptable given that each variable only included a few
items and the low reliability of some variables in the original MSLQ (Pintrich et al., 1991).
2.4.1.4 | Nomological network

The correlations between MSLQ‐C constructs and mathematics achievement were assessed.
2.4.2 | Phase 2: Developing and validating the short form: MSLQ‐CS
2.4.2.1 | Creating the short form

Based on MSLQ‐C, we retained three items for each variable (except for help‐seeking: two items)1 to create the short
form given that three items are the minimum number of items to identify a one‐factor model (Kline, 2015). The
Bhargava‐Ishizuka method (also known as BI‐method) was used to create the short form (Leung & Sachs, 2005). This
method uses the trace information (the sum of all diagonal elements of the variance‐covariance matrix) as the criteria to
select items, which could keep most of the information in the original scale. Compared with factor analytic methods,
which pool all dimensions together, the BI‐method is more appropriate for shortening items within a specific dimension
because it only considers the correlations of items within the specific dimensions. In contrast, items that measure other
dimensions are not considered. More details of the BI‐method can be found in Supplementary Materials.
2.4.2.2 | Construct validity

Set‐ESEM with target rotation was used to investigate the construct validity of the MSLQ‐CS.
2.4.2.3 | Reliability
Coefficient omega (ω) was assessed the reliability of the MSLQ‐CS.
2.4.2.4 | Reproduced information

The amount of reproduced information was assessed by the correlation between the MSLQ‐C and MSLQ‐CS. Levy's
(1967) correction was used to remove the overlapping error variance caused by the same items in MSLQ‐C and MSLQ‐CS.
2.4.2.5 | Nomological network.

The correlations between MSLQ‐CS constructs and mathematics achievement were explored. As the correlation
between MSLQ‐C and mathematics achievement and the correlation between MSLQ‐CS and mathematics
achievement were calculated in Phase 1 and Phase 2, their difference and corresponding significance (measured
by p‐value) were calculated.
1
In the original MSLQ, four items were developed to measure help‐seeking. However, two negatively worded items with low item‐
total correlation were deleted, which resulted in two items for measuring help‐seeking.
10 | WANG ET AL.
3 | RESULTS
3.1 | Phase 1: Adapting and validating the long form: MSLQ‐C
3.1.1 | Initial examination of item performance
Table S2 presents the descriptive statistics of all items. The mean of all items ranged from 4.06 to 9.32, and their
SDs were from 1.36 to 3.41. The items were normally distributed with skewness ranging from −2.51 to 0.32 and
kurtosis ranging from −1.22 to 7.51. Meanwhile, the item‐total correlations of all negatively worded items (i.e., item
33, 37, 40, 57, 52, 60, 77, 80) were lower than the accepted cutoff of 0.30, showing that these items were
inconsistent with other items within the construct (Traub, 1994). Hence, as shown in Table 1, we excluded all
negatively worded items in our subsequent analysis.
3.1.2 | Construct validity
Set‐ESEM was conducted to examine the construct validity of the MSLQ. In this stage, items were excluded when:
(1) items corresponding to the designated factors lower than 0.4 (Hair et al., 2010); (2) items within a set with cross‐
loadings in different factors that were higher than 0.4; (3) high correlation with other items indicated by
modification indices, which means the item was redundant with another item. For example, for item 6 (“I'm certain I
can understand the most difficult course material presented in the readings for the math course”) and item 15 (“I'm
confident I can understand the most complex material presented by the instructor in the math course”), difficult
material and complex material are too similar to be distinguished for students. Hence, item 6 was deleted. The
details of these excluded items can also be found in Table 1.
Finally, the 14‐factor first‐order model demonstrated accepted model fit with RMSEA = 0.039 (90% CI,
0.037–0.042), CFI = 0.917, TLI = 0.902.2 Factor loadings of items on targeted variables ranged from 0.41 to
0.95. Meanwhile, all of the cross‐loadings within sets were smaller than 0.4. The factor loadings can be found in
Table 2.
TABLE 1 Overview of the excluded items
Stage Excluded items Rationale
Initial examination of 33, 37, 52, 57, 77, 80, 60, 40 Reversed‐coded items with low item‐total correlation
item performance
Set‐ESEM 3, 21, 30, 36, 48, 53, 54, 55, Factor loadings < 0.4, cross‐loading items
56, 58, 61, 73, 74, 78
6, 12 High correlation with other items indicated by

modification indices, which means the item was
redundant with other items
Note: Item number is consistent with that in the original MSLQ developed by Pintrich et al. (1991).
2
It should be noted that the variable of effort learning was not identified in this study. In the original version of MSLQ, this variable
has two positively and two negatively worded items. Negative items were reversed before averaging with the positive items.
However, the reverse of the negative may be different from the positive items. Since the number of negatively and positively
worded items is 2 versus 2, no “majority” of items from both sides can dominate the meaning of the factor. This makes the
interpretation of this scale difficult with low reliability. Hence, effort regulation was removed.
TABLE 2 Standardized factor loadings of the Set‐ESEM model for MSLQ‐C
WANG
Affect
Value Expectancy Cognitive and metacognitive strategies Resource management
ET AL.
Items IGO EGO TV CLF SE TA RE ELA ORG CT MSR TE PL HS
1 0.79 −0.01 0.14
16 0.88 0.04 −0.04
22 0.55 0.12 0.18
24 0.65 −0.08 0.04
7 0.00 0.60 0.14
11 ‐0.12 0.70 −0.05
13 0.24 0.54 −0.05
4 0.31 0.05 0.42
10 ‐0.24 0.33 0.59
17 0.01 −0.12 0.91
23 ‐0.11 −0.05 0.93
26 0.11 −0.19 0.82
27 0.19 0.22 0.41
2 0.41 0.25
9 0.68 −0.12
18 0.77 0.08
25 0.85 −0.07
5 0.17 0.61
15 0.01 0.76
20 −0.07 0.83
|
(Continues)
11
12
TABLE 2 (Continued)
|
Affect
29 −0.01 0.86
31 0.00 0.87
8 0.44
14 0.61
19 0.95
28 0.85
39 0.81 −0.09 −0.07 0.10 −0.05
46 0.86 −0.04 −0.01 −0.06 −0.02
59 0.54 0.20 0.04 −0.05 0.14
72 0.48 0.12 0.21 −0.02 0.00
62 0.10 0.66 −0.10 0.10 0.01
64 0.01 0.74 −0.14 0.04 0.19
67 0.15 0.63 0.15 0.01 −0.02
69 −0.01 0.60 0.19 0.07 0.05
81 −0.03 0.60 0.22 0.13 −0.07
32 0.04 0.22 0.69 0.06 −0.12
42 0.00 0.05 0.68 −0.09 0.25
49 0.08 0.12 0.67 0.13 −0.18
63 0.08 −0.20 0.73 0.00 0.24
38 0.08 −0.18 0.08 0.78 −0.13

WANG
ET AL.
TABLE 2 (Continued)
WANG
Affect
ET AL.
47 −0.10 −0.02 0.03 0.88 −0.01
51 0.04 0.09 −0.05 0.70 0.15
66 0.06 0.21 −0.08 0.60 0.17
71 −0.06 0.06 0.00 0.79 0.02
41 0.01 0.01 0.01 0.09 0.72
44 0.08 −0.07 0.11 0.18 0.49
76 0.16 0.05 0.05 0.05 0.49
79 −0.09 0.24 0.13 0.03 0.54
35 0.63 0.06 0.00
43 0.91 0.05 −0.10
65 0.44 0.01 0.08
70 0.55 −0.10 0.15
34 0.21 0.48 0.02
45 −0.08 0.62 0.18
50 −0.02 0.94 −0.08
68 0.00 0.19 0.68
75 0.05 −0.09 0.85
Note: The items that loaded on the factors (>0.4) are presented in bold.
Abbreviations: CLF, control of learning beliefs; CT, critical thinking; EGO, extrinsic goal orientation; ELA, elaboration; HS, help seeking; IGO, intrinsic goal orientation; MSLQ‐C,
Motivated Strategies for Learning Questionnaire‐Chinese version; MSR, metacognitive self‐regulation; ORG, organization; PL, peer learning; RE, rehearsal; SE, self‐efficacy for learning
and performance; Set‐ESEM, set exploratory structural equation modeling; TA, test anxiety; TE, time and study environment management; TV, task value.
| 13
14 | WANG ET AL.
3.1.3 | Reliability
Table 3 shows that the reliabilities of variables on MSLQ‐C were acceptable, ranging from ω = 0.65 (extrinsic goal
orientation) to ω = 0.90 (self‐efficacy for learning and performance) with a mean of ω = 0.80 for all variables.
3.1.4 | Nomological network
Table 3 displays the correlations between mathematics achievement and variables on MSLQ‐C. The correlations
ranged from r = −.01 (text anxiety) to r = .34 (self‐efficacy for learning and performance).
3.2 | Phase 2: Development and validation of the short form: MSLQ‐CS
We retained three items for each variable to create the MSLQ‐CS, which was composed of 41 items in total. The
retained items can be found in Supplementary Materials (see Supplemental questionnaire). Next, we tested whether
MSLQ‐CS displayed a similar factor structure compared with MSLQ‐C. Our results indicated that the MSLQ‐CS
showed an accepted model fit for the 14‐factor first‐order model with RMSEA = 0.033 (90% CI, 0.029 to 0.037),
CFI = 0.949, TLI = 0.933. As shown in Table 4, factor loadings of items on targeted variables ranged from 0.41 to
0.93, and all of the cross‐loadings within sets were smaller than 0.4.
3.2.1 | Reliability
Table 3 shows acceptable reliabilities of MSLQ‐CS, ranging from ω = 0.67 (time and study environment
management) to ω = 0.82 (self‐efficacy for learning and performance) with a mean of ω = 0.75 for all variables.
3.2.2 | Reproduced information
As shown in Table 3, the correlations between the MSLQ‐C and MSLQ‐CS were high, ranging from r = .93 (task
value) to r = .98 (intrinsic goal orientation and organization) with a mean of r = .97. Because the MSLQ‐CS in the
study was embedded in the MSLQ‐C, we then computed Levy's correction correlation that excluded the shared
error variance, which ranged from r = .66 (control of learning beliefs and metacognitive self‐regulation) to r = .78
(self‐efficacy for learning and performance, organization, and critical thinking). The average Levy's corrected
correlation was r = .72, indicating that most information obtained from full scale could be reproduced by the
short form.
3.2.3 | Nomological network
As shown in Table 3, the correlations between mathematics achievement and variables on MSLQ‐CS ranged from
r = −.04 (text anxiety) to r = .34 (time and study environment management). The mean absolute differences in
correlations obtained from the MSLQ‐C and MSLQ‐CS were marginal and nonsignificant, ranging from 0 to 0.07.
These results showed that both MSLQ‐C and MSLQ‐CS displayed similar patterns with mathematics achievement,
further supporting the validity of the MSLQ‐CS.
WANG
ET AL.
TABLE 3 Number of items, reliabilities, reproduced information, and nomological network of MSLQ‐C and MSLQ‐CS
Correlation with mathematics

Number of items Reliabilities Reproduced information achievement
MSLQ‐C MSLQ‐CS MSLQ‐C MSLQ‐CS Without correction Levy's correction MSLQ‐C MSLQ‐CS Δr
Intrinsic goal orientation 4 3 0.74 0.77 0.98 0.75 0.30** 0.28** 0.02
Extrinsic goal orientation 3 3 0.65 0.69 1.00 1.00 0.00 0.00 0.00
Task value 6 3 0.87 0.72 0.93 0.67 0.32** 0.28** 0.04
Control of learning beliefs 4 3 0.79 0.69 0.96 0.66 0.18** 0.19** ‐0.01
Self‐efficacy for learning and performance 5 3 0.90 0.82 0.97 0.78 0.34** 0.33** 0.01
Test anxiety 4 3 0.82 0.74 0.97 0.71 −0.01 ‐0.04 0.03
Rehearsal 4 3 0.80 0.77 0.97 0.73 0.02 0.01 0.01
Elaboration 5 3 0.85 0.74 0.97 0.70 0.24** 0.21** 0.03
Organization 4 3 0.84 0.81 0.98 0.78 0.10* 0.10* 0.00
Critical thinking 5 3 0.89 0.81 0.97 0.78 0.21** 0.18** 0.03
Metacognitive self‐regulation 4 3 0.71 0.70 0.96 0.66 0.28** 0.21** 0.07
Time and study environment management 4 3 0.74 0.67 0.96 0.70 0.31** 0.34** ‐0.03
Peer learning 3 3 0.76 0.76 1.00 1.00 0.15** 0.15** 0.00
Help seeking 2 2 0.77 0.77 1.00 1.00 0.11** 0.11** 0.00
Note: Reproduced information analysis was calculated by the correlations between Motivated Strategies for Learning Questionnaire‐Chinese version (MSLQ‐C) and Motivated Strategies
for Learning Questionnaire‐short form (MSLQ‐CS); Δr (95% confidence interval) is the difference between correlations of self‐regulation variables and mathematics achievement as
obtained from the MSLQ‐C and MSLQ‐CS; For peer learning and help seeking, the number of items for MSLQ‐C and MSLQ‐CS is identical, hence the reliability, reproduced information,
and correlation with mathematics achievement are same.
*p < .05; **p < .01.
| 15
16
TABLE 4 Standardized factor loadings of the Set‐ESEM model for MSLQ‐CS

|
Affect
16 0.77 0.03 0.04
22 0.63 0.09 0.15
24 0.72 −0.03 −0.04
7 0.06 0.60 0.10
11 −0.08 0.82 −0.10
13 0.18 0.49 0.05
10 −0.30 0.11 0.76
17 0.24 −0.17 0.67
27 0.14 0.06 0.53
2 0.57 0.12
9 0.68 −0.15
18 0.70 0.07
5 0.14 0.64
15 −0.08 0.81
31 0.00 0.86
8 0.58
14 0.78
19 0.73
39 0.83 −0.04 −0.12 0.06 0.02
46 0.82 −0.04 0.02 −0.07 0.04
72 0.41 0.13 0.28 −0.01 −0.02

WANG
62 0.09 0.61 −0.09 0.07 0.13

ET AL.
TABLE 4 (Continued)
WANG
Affect
ET AL.
69 −0.02 0.41 0.28 0.18 0.08
81 0.02 0.79 0.06 0.05 −0.01
32 0.16 0.38 0.47 0.06 −0.13
42 −0.03 0.05 0.78 −0.04 0.13
63 0.08 −0.17 0.78 0.06 0.08
38 0.09 −0.12 0.04 0.78 −0.14
47 −0.09 0.01 −0.02 0.90 0.00
66 0.00 0.23 −0.05 0.49 0.26
41 −0.02 0.03 −0.02 0.00 0.81
44 0.08 0.02 0.03 0.09 0.58
76 0.07 0.04 0.16 0.00 0.49
43 0.87 0.08 −0.10
65 0.42 0.02 0.08
70 0.56 −0.10 0.15
34 0.22 0.49 0.01
45 −0.08 0.62 0.18
50 −0.03 0.93 −0.07
68 0.00 0.19 0.67
75 0.05 −0.09 0.86
Note: The items that loaded on the factors (>0.4) are presented in bold.
Abbreviations: CLF, control of learning beliefs; CT, critical thinking; EGO, extrinsic goal orientation; ELA, elaboration; HS, help seeking; IGO, intrinsic goal orientation; MSLQ‐CS,
|
Motivated Strategies for Learning Questionnaire‐short form; MSR, metacognitive self‐regulation; ORG, organization; PL, peer learning; RE, rehearsal; SE, self‐efficacy for learning and
performance; Set‐ESEM, set exploratory structural equation modeling; TA, test anxiety; TE, time and study environment management; TV, task value.
17
18 | WANG ET AL.
4 | DISC US SION
The MSLQ has been extensively used to explore the role of self‐regulated learning in academic settings. To
generalize the application of the MSLQ to the Chinese context and the domain of mathematics, the first goal of this
study was to adapt and validate the MSLQ for Chinese senior secondary students in mathematics learning (i.e.,
MSLQ‐C). The second goal was to develop the short form of MSLQ that can be used to measure students' self‐
regulations when short measures are necessary (i.e., MSLQ‐CS). The results demonstrated satisfactory
psychometric properties for both MSLQ‐C and MSLQ‐CS.
For construct validity, both MSLQ‐C and MSLQ‐CS achieved an acceptable fit for the 14‐factor correlated
model, which corroborates the original factor structure proposed by Pintrich et al. (1993). This study extends the
existing research by exploring the factor structure of MSLQ with Set‐ESEM. Compared with EFA, CFA, and ESEM,
Set‐ESEM considers the relationship across and within each domain, making the factor structure more rigorous
(Marsh et al., 2020).
Acceptable reliability was found for each variable, which is consistent with previous research (e.g., Pintrich
et al., 1993). This finding suggests that both MSLQ‐C and MSLQ‐CS possess acceptable internal consistencies in the
Chinese context. It is not surprising that the reliabilities of the MSLQ‐CS were slightly lower than MSLQ‐C as true
variance increases more rapidly than error variance when the number of items increases (Embretson & Reise, 2000).
For example, Kruyen et al. (2013) analyzed the 137 pair of long scales and their corresponding short form, and they
found the mean reliabilities in internal consistency decreased from α = .84–.77. However, intrinsic and extrinsic goal
orientation are two exceptions because the short form' reliabilities were marginally higher than full scale,
presumably because problematic items (i.e., items with correlated errors) were deleted during the development of
the short form.
We found that all variables, except for extrinsic goal orientation, test anxiety, and rehearsal, had significant and
low‐to‐medium correlations with mathematics achievement. These findings are in accordance with the evidence
reported in previous studies (e.g., Ocak & Yamac, 2013; Shores & Shannon, 2007), which revealed the applicability
of MSLQ‐C and MSLQ‐CS to be associated with academic achievement. Furthermore, both MSLQ‐C and MSLQ‐CS
showed similar correlations with mathematics achievement with marginal and nonsignificant differences. Hence,
our results demonstrated that MSLQ‐C and MSLQ‐CS are embedded in a similar nomological network and could
help researchers draw similar results when connecting it with mathematics achievement.
Our study also indicated that the MSLQ‐C and MSLQ‐CS were highly correlated, which showed that a large
amount of information measured by the MSLQ‐C could be reproduced by MSLQ‐CS (Girard & Christensen, 2008;
Levy, 1967). This implies that the items in the short form are the most representative of the full scale and could
assess the complexity of the theoretical construct of self‐regulated learning.
Compared with previous research that developed short forms by removing variables (e.g., Pintrich & de
Groot, 1990), our study selected subsets of items within each variable. This method extends previous research by
keeping almost all variables within the self‐regulated theoretical framework proposed by Pintrich and Zusho (2002).
With MSLQ‐CS, researchers could explore the relationship between self‐regulated learning behaviors with less
burden placed on students than the MSLQ‐C.
5 | C ON T R I B U T I O N S , L I M I T A T I O N S , A ND F U T U R E D I R E C T I O N S
The present study makes three significant contributions. First, we adapted the MSLQ for Chinese secondary
students in mathematics learning. The adapted questionnaire extends the application scope of the MSLQ to
measure Chinese students' mathematics learning, which could help researchers extend self‐regulated learning
research in the Chinese context.
WANG ET AL. | 19
Second, the MSLQ‐CS exhibited acceptable psychometric properties, which provide an alternative for
researchers to capture the students' self‐regulation with fewer items. The MSLQ‐CS with fewer items is practical
and user‐friendly. It is easier to be administered due to the fewer constraints of administration time and less space
in the overall survey package. Furthermore, it might improve the quality of research data because it could reduce
participants' fatigue, frustration, and boredom brought by long response time and redundant items.
Third, the present study would help measure and potentiall improve Chinese secondary students' mathematics
learning. This study has demonstrated the close relationship between self‐regulated learning and mathematics
achievement (Dent & Koenka, 2016). Using this psychometrically sound instrument, researchers and educators
could better understand the current status of students' self‐regulated learning and develop intervention targets to
help them develop a high level of self‐regulation in mathematics learning.
Despite the above‐mentioned contributions, a few limitations need to be addressed in future studies. First, negatively
worded items were removed from the original MSLQ according to the results of item performance, which is consistent
with previous research (Law et al., 2008; Rao & Sachs, 1999; Sachs et al., 2001). Credé et al. (2009) argued that positively
and negatively worded items might measure different constructs. Many studies on MSLQ have indicated that Chinese
students had some problems responding to the negatively worded items, which were clustered to form a separate
“method” scale and removed (Law et al., 2008; Rao & Sachs, 1999; Sachs et al., 2001). However, some researchers may
argue that negatively worded items are essential to control for acquiescence response style (Nunnally & Bernstein, 1994),
and a balance of positively and negatively worded items might increase the validity of the measures (Kam & Meyer, 2015).
Hence, more advanced statistical methods (e.g., factor mixture modeling; Kam & Fan, 2020) are recommended in future
studies to explore the role of negatively worded items in the MSLQ.
Second, the data were collected at the same time point. Hence, this study was cross‐sectional in nature. We did
not conduct the predictive validity given that the best way to establish predictive validity is through conducting a
longitudinal study. We encourage future studies to adopt a more rigorous research design by collecting longitudinal
data to examine the predictive validity of the MSLQ. Furthermore, the test‐retest reliability information obtained
from repeated assessments would also be useful.
Third, in line with previous research (e.g., Gogol et al., 2014; Sánchez et al., 2021), our short form was administrated as
part of the corresponding full scale. This may overestimate the relationship between short form and full scale because of
the shared correlated measurement error. Although Levy's correction (1967) used in our study accounted for shared error,
it would be advisable to assess agreement between the full scale and the short form with independent samples.
Fourth, the participants of this study were all drawn from Guangdong province, which is more economically
advanced than other provinces in China. We encourage future studies to replicate our research in other regions of
China to examine the generalizability of the findings.
Fifth, we deleted the redundant items mainly based on the psychometric properties of the items (i.e., factor
loading, modification indices, and item‐total correlation). This kind of data‐driven decision might ignore the
differences in the content. Hence, we encourage future studies to delete the redundant items with a combination of
data‐driven and theory‐driven approaches. For example, a qualitative interview can be conducted to explore
students' nuanced understanding of the similarities and differences in the redundant items.
6 | C ONC LUS I ON
A reliable and valid instrument is an essential step to effective measurement, intervention, and improvement. This
study adapted and validated the MSLQ for Chinese secondary students in mathematics learning. The MSLQ‐C
demonstrated acceptable construct validity, reliability, and concurrent validity. More importantly, we developed a
short form, the MSLQ‐CS, which is psychometrically sound and able to replicate the information obtained from the
longer form of the MSLQ‐C. The short form can reduce participants' burden and be used when the test time and
space are limited, providing a practical tool for mathematics education researchers.
20 | WANG ET AL.
CO NFL I CTS OF I NTEREST

The authors declare no conflicts of interest.
D A TA A V A I L A B I L I T Y S T A T E M E N T
The data that support the findings of this study are available from the corresponding author upon reasonable
request.
ORCID
Faming Wang http://orcid.org/0000-0001-7144-2939
REFERENCES
Bonanomi, A., Olivari, M. G., Mascheroni, E., Gatti, E., & Confalonieri, E. (2018). Using a multidimensional Rasch analysis to
evaluate the psychometric properties of the Motivated Strategies for Learning Questionnaire (MSLQ) among high
school students. Testing, Psychometrics, Methodology in Applied Psychology, 25, 83–100. https://doi.org/10.4473/
TPM25.1.5
Cai, Y., King, R. B., & McInerney, D. M. (2022). The concurrent trajectories of utility value, metacognitive strategy use, and
achievement. The Journal of Experimental Education, 1–22. https://doi.org/10.1080/00220973.2022.2053496
Cho, E., & Kim, S. (2015). Cronbach's coefficient alpha: Well known but poorly understood. Organizational Research
Methods, 18(2), 207–230. https://doi.org/10.1177/1094428114555994
Christopher Westland, J. (2010). Lower bounds on sample size in structural equation modeling. Electronic Commerce
Research and Applications, 9(6), 476–487. https://doi.org/10.1016/j.elerap.2010.07.003
Cleary, T. J., Slemp, J., & Pawlo, E. R. (2021). Linking student self‐regulated learning profiles to achievement and
engagement in mathematics. Psychology in the Schools, 58(3), 443–457. https://doi.org/10.1002/pits.22456
Cleary, T. J., & Zimmerman, B. J. (2004). Self‐regulation empowerment program: A school‐based program to enhance self‐
regulated and self‐motivated cycles of student learning. Psychology in the Schools, 41(5), 537–550. https://doi.org/10.
1002/pits.10177
Cohen, J. (1988). Statistical power analysis for the behavioral sciences (2nd ed.). Lawrence Earlbaum Associates.
Credé, M., Chernyshenko, O. S., Bagraim, J., & Sully, M. (2009). Contextual performance and the job satisfaction–dissatisfaction
distinction: Examining artifacts and utility. Human Performance, 22(3), 246–272. https://doi.org/10.1080/08959
280902970427
Credé, M., Harms, P., Niehorster, S., & Gaye‐Valentine, A. (2012). An evaluation of the consequences of using short
measures of the Big Five personality traits. Journal of Personality and Social Psychology, 102(4), 874–888. https://doi.
org/10.1037/a0027403
Credé, M., & Phillips, L. A. (2011). A meta‐analytic review of the motivated strategies for learning questionnaire. Learning
and Individual Differences, 21(4), 337–346. https://doi.org/10.1016/j.lindif.2011.03.002
Dent, A. L., & Koenka, A. C. (2016). The relation between self‐regulated learning and academic achievement across
childhood and adolescence: A meta‐analysis. Educational Psychology Review, 28(3), 425–474. https://doi.org/10.
1007/s10648-015-9320-8
Dicke, T., Marsh, H. W., Riley, P., Parker, P. D., Guo, J., & Horwood, M. (2018). Validating the Copenhagen Psychosocial
Questionnaire (COPSOQ‐II) using set‐ESEM: Identifying psychosocial risk factors in a sample of school principals.
Frontiers in Psychology, 9, 584. https://doi.org/10.3389/fpsyg.2018.00584
DiGiacomo, G., & Chen, P. P. (2016). Enhancing self‐regulatory skills through an intervention embedded in a middle school
mathematics curriculum. Psychology in the Schools, 53(6), 601–616. https://doi.org/10.1002/pits.21929
Duncan, T. G., & McKeachie, W. J. (2005). The making of the motivated strategies for learning questionnaire. Educational
Psychologist, 40(2), 117–128. https://doi.org/10.1207/s15326985ep4002_6
Dunn, T. J., Baguley, T., & Brunsden, V. (2014). From alpha to omega: A practical solution to the pervasive problem of
internal consistency estimation. British Journal of Psychology, 105(3), 399–412. https://doi.org/10.1111/bjop.12046
Embretson, S. E., & Reise, S. P. (2000). Item response theory for psychologists. Erlbaum.
Enders, C. K. (2010). Applied missing data analysis (Methodology in the social sciences). Guilford Publications.
Girard, T. A., & Christensen, B. K. (2008). Clarifying problems and offering solutions for correlated error when assessing the
validity of selected‐subtest short forms. Psychological Assessment, 20(1), 76–80. https://doi.org/10.1037/1040-3590.
20.1.76
Gogol, K., Brunner, M., Goetz, T., Martin, R., Ugen, S., Keller, U., Fischbach, A., & Preckel, F. (2014). My questionnaire is too
long!” The assessments of motivational‐affective constructs with three‐item and single‐item measures. Contemporary
Educational Psychology, 39(3), 188–205. https://doi.org/10.1016/j.cedpsych.2014.04.002
WANG ET AL. | 21
Graham, J. M. (2006). Congeneric and (essentially) tau‐equivalent estimates of score reliability: What they are and how to use
them. Educational and Psychological Measurement, 66(6), 930–944. https://doi.org/10.1177/0013164406288165
Gravemeijer, K., Stephan, M., Julie, C., Lin, F. L., & Ohtani, M. (2017). What mathematics education may prepare students
for the society of the future? International Journal of Science and Mathematics Education, 15(1), 105–123. https://doi.
org/10.1007/s10763-017-9814-6
Hair, J. F., Black, W. C., Babin, B., & Anderson, R. E. (2010). Multivariate data analysis (7th ed.). Prentice Hall.
Hau, K. T., & Ho, I. T. (2010). Chinese students' motivation and achievement. In M. H. Bond (Ed.), The Oxford handbook of
Chinese psychology (pp. 187–204). Oxford University Press.
Haw, J. Y., & King, R. B. (2022). Need‐supportive teaching is associated with reading achievement via intrinsic motivation
across eight cultures. Learning and Individual Differences, 97, 102161. https://doi.org/10.1016/j.lindif.2022.102161
Hilpert, J. C., Stempien, J., van der Hoeven Kraft, K. J., & Husman, J. (2013). Evidence for the latent factor structure of the
MSLQ: A new conceptualization of an established questionnaire. SAGE Open, 3(4), 215824401351030. https://doi.
org/10.1177/2158244013510305
Hwang, T. M. (2015). The studying and striving of secondary students. In S. Hsu, & Y.‐Y. Wu (Eds.), Education as cultivation
in Chinese culture (pp. 127–148). Springer Science.
International Test Commission. (2018). ITC guidelines for translating and adapting tests. International Journal of Testing,
18(2), 101–134. https://doi.org/10.1080/15305058.2017.1398166
Jackson, C. R. (2018). Validating and adapting the Motivated Strategies for Learning Questionnaire (MSLQ) for STEM
courses at an HBCU. AERA Open, 4(4), 233285841880934. https://doi.org/10.1177/2332858418809346
Kam, C. C. S., & Fan, X. (2020). Investigating response heterogeneity in the context of positively and negatively worded
items by using factor mixture modeling. Organizational Research Methods, 23(2), 322–341. https://doi.org/10.1177/
1094428118790371
Kam, C. C. S., & Meyer, J. P. (2015). How careless responding and acquiescence response bias can influence construct
dimensionality: The case of job satisfaction. Organizational Research Methods, 18, 512–541. https://doi.org/10.1177/
1094428115571894
Kennedy, K., & Lee, J. C. K. (2007). The changing role of schools in Asian societies: Schools for the knowledge society. Routledge.
King, R. B. (2022). Sociocultural and ecological perspectives on achievement motivation. Asian Journal of Social Psychology,
25(3), 433–448. https://doi.org/10.1111/ajsp.12507
King, R. B., & McInerney, D. M. (2014). Culture's consequences on student motivation: Capturing cross‐cultural universality
and variability through personal investment theory. Educational Psychologist, 49(3), 175–198. https://doi.org/10.
1080/00461520.2014.926813
King, R. B., & McInerney, D. M. (2016). Culturalizing motivation research in educational psychology. British Journal of
Educational Psychology, 86(1), 1–7. https://doi.org/10.1111/bjep.12106
Kline, P. (2000). The handbook of psychological testing (2nd ed.). Routledge.
Kline, R. B. (2005). Principles and practice of structural equation modeling (2nd ed.). Guilford Press.
Kruyen, P. M., Emons, W. H. M., & Sijtsma, K. (2013). On the shortcomings of shortened tests: A literature review.
International Journal of Testing, 13, 223–248. https://doi.org/10.1080/15305058.2012.703734
Law, Y., Chan, C. K. K., & Sachs, J. (2008). Beliefs about learning, self‐regulated strategies and text comprehension
among Chinese children. British Journal of Educational Psychology, 78(1), 51–73. https://doi.org/10.1348/00070
9907X179812
Lee, J. C., Yin, H., & Zhang, Z. (2010). Adaptation and analysis of Motivated Strategies for Learning Questionnaire in the
Chinese setting. International Journal of Testing, 10(2), 149–165. https://doi.org/10.1080/15305050903534670
Leung, F. K. S. (2001). In search of an East Asian identity in mathematics education. Educational Studies in Mathematics,
47(1), 35–51. https://doi.org/10.1023/A:1017936429620
Leung, S. O. (2011). A comparison of psychometric properties and normality in 4‐, 5‐, 6‐, and 11‐point Likert scales. Journal
of Social Service Research, 37(4), 412–421. https://doi.org/10.1080/01488376.2011.580697
Leung, S. O., & Sachs, J. (2005). Bhargava and ishizuka's BI‐method: A neglected method for variable selection. The Journal
of Experimental Education, 73(4), 353–367. https://doi.org/10.3200/JEXE.73.4.353-367
Levy, P. (1967). The correction for spurious correlation in the evaluation of short‐form tests. Journal of Clinical Psychology,
23(1), 84–86. https://doi.org/10.1002/1097-4679(196701)23:1>84::AID-JCLP2270230123<3.0.CO;2-2
Li, J. (2003). US and Chinese cultural beliefs about learning. Journal of Educational Psychology, 95, 258–267. https://doi.org/
10.1037/0022-0663.95.2.258
Lucke, J. F. (2005). The α and the ω of congeneric test theory: An extension of reliability and internal consistency to
heterogeneous tests. Applied Psychological Measurement, 29(1), 65–81. https://doi.org/10.1177/0146621604272739
Marsh, H. W., Guo, J., Dicke, T., Parker, P. D., & Craven, R. G. (2020). Confirmatory factor analysis (CFA), exploratory
structural equation modeling (ESEM), and set‐ESEM: Optimal balance between goodness of fit and parsimony.
Multivariate Behavioral Research, 55(1), 102–119. https://doi.org/10.1080/00273171.2019.1602503
22 | WANG ET AL.
Marsh, H. W., Hau, K.‐T., & Wen, Z. (2004). In search of golden rules: Comment on hypothesis‐testing approaches to setting
cutoff values for fit indexes and dangers in overgeneralizing Hu and Bentler's (1999) findings. Structural Equation
Modeling: A Multidisciplinary Journal, 11, 320–341. https://doi.org/10.1207/s15328007sem1103_2
Mclnerney, D. M., & King, R. B. (2018). Culture and self‐regulation in educational contexts. In D. H. Schunk, & J. A. Greene (Eds.),
Educational psychology handbook series. Handbook of self‐regulation of learning and performance (pp. 485–502). Routledge.
Meijs, C., Neroni, J., Gijselaers, H. J. M., Leontjevas, R., Kirschner, P. A., & de Groot, R. H. M. (2019). Motivated strategies for
learning questionnaire part B revisited: New subscales for an adult distance education setting. The Internet and Higher
Education, 40, 1–11. https://doi.org/10.1016/j.iheduc.2018.09.003
Morin, A. J. S., Marsh, H. W., & Nagengast, B. (2013). Exploratory structural equation modeling. In G. R. Hancock, & R. O.
Mueller (Eds.), Structural equation modeling: A second course (2nd ed., pp. 395–436). IAP.
Mullis, I. V. S., Martin, M. O., Foy, P., Kelly, D. L., & Fishbein, B. (2020). TIMSS 2019 international results in mathematics and
science. TIMSS & PIRLS International Study Center. https://timssandpirls.bc.edu/timss2019/international-results/
Muthén, L. K., & Muthén, B. O. (2017). Mplus user's guide. Muthén &Muthén.
Nunnally, J. C., & Bernstein, I. H. (1994). Psychometric theory (3rd ed.). McGraw‐Hill.
Ocak, G., & Yamac, A. (2013). Examination of the relationships between fifth graders' self‐regulated learning strategies,
motivational beliefs, attitudes, and achievement. Educational Sciences: Theory & Practice, 13(1), 380–387.
OECD. (2020). PISA 2018 technical report. OECD Publishing. https://www.oecd.org/pisa/data/pisa2018technicalreport/
Panadero, E. (2017). A review of self‐regulated learning: Six models and four directions for research. Frontiers in Psychology,
8, 422. https://doi.org/10.3389/fpsyg.2017.00422
Perels, F., Dignath, C., & Schmitz, B. (2009). Is it possible to improve mathematical achievement by means of self‐regulation
strategies? Evaluation of an intervention in regular math classes. European Journal of Psychology of Education, 24(1),
17–31. https://doi.org/10.1007/BF03173472
Pintrich, P., Wolters, C., & Baxter, G. (2000). Assessing metacognition and self‐regulated learning. In G. Schraw, & J. Impara
(Eds.), Issues in the measurement of cognition (pp. 44–97). Buros Institute of Mental Measurement.
Pintrich, P. R. (2000). The role of goal orientation in self-regulated learning. In M. Boekaerts, P. R. Pintrich & M. Zeidner
(Eds.), Handbook of self‐regulation (pp. 451–502). San Diego, CA: Academic.
Pintrich, P. R. (2004). A conceptual framework for assessing motivation and self‐regulated learning in college students.
Educational Psychology Review, 16(4), 385–407. https://doi.org/10.1007/s10648-004-0006-x
Pintrich, P. R., & de Groot, E. V. (1990). Motivational and self‐regulated learning components of classroom academic
performance. Journal of Educational Psychology, 82(1), 33–40. https://doi.org/10.1037/0022-0663.82.1.33
Pintrich, P. R., Smith, D. A. F., García, T., & McKeachie, W. J. (1991). A manual for the use of the Motivated Strategies for
Learning Questionnaire (MSLQ). University of Michigan.
Pintrich, P. R., Smith, D. A. F., Garcia, T., & McKeachie, W. J. (1993). Reliability and predictive validity of the Motivated
Strategies for Learning Questionnaire (MSLQ). Educational and Psychological Measurement, 53, 801–813. https://doi.
org/10.1177/0013164493053003024
Pintrich, P. R., & Zusho, A. (2002). The development of academic self‐regulation: The role of cognitive and motivational
factors. In A. Wigfield & J. S. Eccles (Eds.), Development of achievement motivation (pp. 249–284). Academic.
Pratt, D. D., Kelly, M., & Wong, W. S. S. (1999). Chinese conceptions of “effective teaching” in Hong Kong: Towards
culturally sensitive evaluation of teaching. International Journal of Lifelong Education, 18, 241–258. https://doi.org/10.
1080/026013799293739a
Rao, N., & Sachs, J. (1999). Confirmatory factor analysis of the Chinese version of the motivated strategies for
learning questionnaire. Educational and Psychological Measurement, 59(6), 1016–1029. https://doi.org/10.1177/
00131649921970206
Robson, D. A., Allen, M. S., & Howard, S. J. (2020). Self‐regulation in childhood as a predictor of future outcomes: A meta‐
analytic review. Psychological Bulletin, 146(4), 324–354. https://doi.org/10.1037/bul0000227
Sachs, J., Law, Y., Chan, C. K. K., & Rao, N. (2001). A nonparametric item analysis of the motivated strategies for learning
questionnaire—Chinese version. Psychologia, 44, 197–208. https://doi.org/10.2117/psysoc.2001.197
Sánchez, T., León, J., Gilar‐Corbi, R., & Castejón, J. L. (2021). Validation of a short scale for student evaluation of teaching
ratings in a polytechnic higher education institution. Frontiers in Psychology, 12, Article 635543. https://doi.org/10.
3389/fpsyg.2021.635543
Schunk, D. H. (2005). Self‐regulated learning: The educational legacy of Paul R. Pintrich. Educational Psychologist, 40(2),
85–94. https://doi.org/10.1207/s15326985ep4002_3
Shores, M. L., & Shannon, D. M. (2007). The effects of self‐regulation, motivation, anxiety, and attributions on mathematics
achievement for fifth and sixth grade students. School Science and Mathematics, 107(6), 225–236. https://doi.org/10.
1111/j.1949-8594.2007.tb18284.x
Soper, D. S. (2022). A‐priori sample size calculator for structural equation models [Software]. https://www.danielsoper.
com/statcalc
WANG ET AL. | 23
Thomas, C. L., & Cassady, J. C. (2019). The influence of personality factors, value appraisals, and control appraisals on
cognitive test anxiety. Psychology in the Schools, 56(10), 1568–1582. https://doi.org/10.1002/pits.22303
Tong, F., Guo, H., Wang, Z., Min, Y., Guo, W., & Yoon, M. (2019). Examining cross‐cultural transferability of self‐regulated
learning model: An adaptation of the Motivated Strategies for Learning Questionnaire for Chinese adult learners.
Educational Studies, 1–18. https://doi.org/10.1080/03055698.2019.1590183
Traub, R. E. (1994). Reliability for the social sciences: Theory and applications. Sage.
Walker, C. O., Greene, B. A., & Mansell, R. A. (2006). Identification with academics, intrinsic/extrinsic motivation, and self‐
efficacy as predictors of cognitive engagement. Learning and Individual Differences, 16(1), 1–12. https://doi.org/10.
1016/j.lindif.2005.06.004
Wang, F., Leung, S. O., & Jiang, C. (2021). Psychometric properties of cognitive and metacognitive learning strategies scale
among Chinese senior secondary school students. Journal of Psychoeducational Assessment, 39(6), 761–771. https://
doi.org/10.1177/07342829211011806
Wang, J., King, R. B., & Rao, N. (2019). The role of social‐academic goals in Chinese students' self‐regulated learning.
European Journal of Psychology of Education, 34(3), 579–600. https://doi.org/10.1007/s10212-018-0404-y
Wang, T., & Cai, J. (2007). Chinese (Mainland) teachers' views of effective mathematics teaching and learning. ZDM, 39(4),
287–300. https://doi.org/10.1007/s11858-007-0030-7
Wolters, C. A., & Pintrich, P. R. (1998). Contextual differences in student motivation and self‐regulated learning in
mathematics, English, and social studies classrooms. Instructional Science, 26, 27–47. https://doi.org/10.1023/
A:1003035929216
Wolters, C. A., Yu, S. L., & Pintrich, P. R. (1996). The relation between goal orientation and students' motivational beliefs
and self‐regulated learning. Learning and individual differences, 8(3), 211–238. https://doi.org/10.1016/S1041-
6080(96)90015-1
Wong, L. H., Chai, C. S., Chen, W., & Chin, C. K. (2013). Measuring Singaporean students' motivation and strategies of bilingual
learning. The Asia‐Pacific Education Researcher, 22(3), 263–272. https://doi.org/10.1007/s40299-012-0032-2
Xu, J. (2020). High level cognitive strategies scale for middle school students: A psychometric evaluation. Current
Psychology, 41, 2711–2718. https://doi.org/10.1007/s12144-020-00786-1
Yukhymenko‐Lescroart, M. A., & Gilbert, W. D. (2021). Development and validation of the Coaching Athlete Purpose Scale
(CAPS). Psychology of Sport and Exercise, 56, Article 101960. https://doi.org/10.1016/j.psychsport.2021.101960
Zhang, D., Li, S., & Tang, R. (2004). The “two basics:” Mathematics teaching and learning in Mainland China. In L. Fan, N.‐Y.
Wong, J. Cai, & S. Li (Eds.), How Chinese learn mathematics: Perspectives from insiders (pp. 189–207). World Scientific.
Zhao, W. (2020). Predicament and outlook of China's math education. National Science Review, 7(9), 1513–1517. https://
doi.org/10.1093/nsr/nwaa070
Zhou, Y., & Wang, J. (2021). Psychometric properties of the MSLQ‐B for adult distance education in China. Frontiers in
Psychology, 12, 441. https://doi.org/10.3389/fpsyg.2021.620564
Ziegler, M., Kemper, C. J., & Kruyen, P. (2014). Short scales – Five misunderstandings and ways to overcome them. Journal
of Individual Differences, 35, 185–189. https://doi.org/10.1027/1614-0001/a000148
Zilberman, A., & Ice, L. (2021). Why computer occupations are behind strong STEM employment growth in the 2019‐29
decade. Beyond the Numbers: Employment & Unemployment, 10(1). https://www.bls.gov/opub/btn/volume-10/why-
computer-occupationsare-behind-strong-stem-employment-growth.htm
S UP P O R T I N G I N F O R M A T I O N
Additional supporting information can be found online in the Supporting Information section at the end of this
article.
How to cite this article: Wang, F., Jiang, C., King, R. B., & Leung, S. O. (2022). Motivated Strategies for
Learning Questionnaire (MSLQ): Adaptation, validation, and development of a short form in the Chinese
context for mathematics. Psychology in the Schools, 1–23. https://doi.org/10.1002/pits.22845

Motivated Strategies For Learning Questionnaire MSLQ

Uploaded by

Copyright:

Available Formats

Motivated Strategies For Learning Questionnaire MSLQ

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Motivated Strategies For Learning Questionnaire MSLQ

Uploaded by

Copyright:

Available Formats

15206807, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/pits.22845 by University Of Macau Wu Yee Sun Library (E2), Wiley Online Library on [30/12/2022].

Motivated Strategies for Learning Questionnaire

Faming Wang1 | Chunlian Jiang2 | Ronnel B. King3 |

Psychology in the Schools. 2022;1–23. wileyonlinelibrary.com/journal/pits © 2022 Wiley Periodicals LLC. | 1

1.1 | Self‐regulated learning and mathematics learning

1.3 | Psychometric properties of MSLQ

1.4 | The need for a short form of MSLQ

1.5 | Set‐exploratory structural equation modeling (Set‐ESEM)

1.6 | The Chinese context

1.7 | The present study

2.2 | Procedures for translating MSLQ

2.3.2 | Mathematics achievement

2.4 | Data analyses

2.4.1 | Phase 1: Adapting and validating the long form: MSLQ‐C

2.4.1.1 | Initial examination of item performance

2.4.1.2 | Construct validity

2.4.1.4 | Nomological network

2.4.2 | Phase 2: Developing and validating the short form: MSLQ‐CS

2.4.2.1 | Creating the short form

2.4.2.2 | Construct validity

2.4.2.4 | Reproduced information

2.4.2.5 | Nomological network.

3.1 | Phase 1: Adapting and validating the long form: MSLQ‐C

3.1.1 | Initial examination of item performance

3.1.2 | Construct validity

TABLE 1 Overview of the excluded items

Stage Excluded items Rationale

6, 12 High correlation with other items indicated by

Items IGO EGO TV CLF SE TA RE ELA ORG CT MSR TE PL HS

1 0.79 −0.01 0.14

16 0.88 0.04 −0.04

22 0.55 0.12 0.18

24 0.65 −0.08 0.04

7 0.00 0.60 0.14

11 ‐0.12 0.70 −0.05

13 0.24 0.54 −0.05

4 0.31 0.05 0.42

10 ‐0.24 0.33 0.59

17 0.01 −0.12 0.91

23 ‐0.11 −0.05 0.93

26 0.11 −0.19 0.82

27 0.19 0.22 0.41

39 0.81 −0.09 −0.07 0.10 −0.05

46 0.86 −0.04 −0.01 −0.06 −0.02

59 0.54 0.20 0.04 −0.05 0.14

72 0.48 0.12 0.21 −0.02 0.00

62 0.10 0.66 −0.10 0.10 0.01

64 0.01 0.74 −0.14 0.04 0.19

67 0.15 0.63 0.15 0.01 −0.02

69 −0.01 0.60 0.19 0.07 0.05

81 −0.03 0.60 0.22 0.13 −0.07

32 0.04 0.22 0.69 0.06 −0.12

42 0.00 0.05 0.68 −0.09 0.25

49 0.08 0.12 0.67 0.13 −0.18

63 0.08 −0.20 0.73 0.00 0.24

38 0.08 −0.18 0.08 0.78 −0.13

Items IGO EGO TV CLF SE TA RE ELA ORG CT MSR TE PL HS

47 −0.10 −0.02 0.03 0.88 −0.01

Task value 6 3 0.87 0.72 0.93 0.67 0.32 0.28 0.04

Elaboration 5 3 0.85 0.74 0.97 0.70 0.24 0.21 0.03

Critical thinking 5 3 0.89 0.81 0.97 0.78 0.21 0.18 0.03

Metacognitive self‐regulation 4 3 0.71 0.70 0.96 0.66 0.28 0.21 0.07

Peer learning 3 3 0.76 0.76 1.00 1.00 0.15 0.15 0.00

Help seeking 2 2 0.77 0.77 1.00 1.00 0.11 0.11 0.00