Revisiting the Motivated Strategies for Learning Questionnaire

Abstract

The Motivated Strategies for Learning Questionnaire (MSLQ) has dominated self-regulated learning research since the early 1990s. In this study, the two MSLQ subscales specifically designed to assess self-regulation—Metacognitive Self-Regulation subscale and Effort Regulation subscale—were examined. Results indicated that the structure of the two scales is not supported by the original data reported by Pintrich, Smith, Garcia, and McKeachie in 1991 or new data. Statistical and theoretical analyses supported two modified scales, the General Strategies for Learning scale and the Clarification Strategies for Learning scale, that assess academic self-regulation from the original MSLQ items. The statistical and theoretical analyses, results, and modified scales are discussed.

Keywords

factor analysis motivation self-regulation

Numerous aspects of motivation and cognition affect academic success; one important facet of learning that is associated with positive student outcomes is self-regulation (Pintrich & DeGroot, 1990; Zimmerman, 1990). Self-regulated learning “occurs largely from the influence of students’ self-generated thoughts, feelings, strategies, and behaviors, which are oriented toward the attainment of goals” (Schunk & Zimmerman, 1998, p. viii). Recent research links self-regulation to prefrontal cortex functioning and a number of social and mental health disorders (Heatherton & Wagner, 2011), including attention deficit hyperactivity disorder (Barkley, 2010). These findings indicate that self-regulation is not simply something students need to be successful in school; instead, self-regulatory skills also facilitate happiness and success in life beyond the classroom.

In light of the connection of self-regulation to academic success, to mental and social health, and to the escalating numbers of attention deficit hyperactivity disorder diagnoses, it is heartening to note that research continues to support that targeted instruction leads to improved self-regulatory abilities (e.g., Greene & Azevedo, 2007; Paris & Winograd, 2001; Perels, Dignath, & Schmitz, 2009; Perels, Gurtler, & Schmitz, 2005; Stoeger & Ziegler, 2008). Thus, the measurement of students’ aptitude for self-regulation is of paramount importance if one wants to address and improve learners’ self-regulation, and subsequently, academic achievement (Boekaerts & Corno, 2005). The purpose of this research is to address theoretical and statistical issues that exist with the Motivated Strategies for Learning Questionnaire (MSLQ).

In the research literature regarding self-regulated learning, the MSLQ (Pintrich, Smith, Garcia, & McKeachie, 1993) is one of the most widely used measures (Duncan & McKeachie, 2005; Winne & Perry, 2000). In fact, the Social Sciences Citation Index reports that the article in which the MSLQ was first introduced has been cited more than 300 times. The MSLQ was developed from a social–cognitive theoretical perspective, and it consists of two primary scales: Motivation Scale and Learning Strategies Scale. The Motivation Scale contains six subscales with 31 items regarding students’ goals, beliefs, skills, and anxiety related to their courses and tests. The Learning Strategies Scale includes nine subscales with 50 items assessing cognitive strategies and resources management skills. The 15 subscales use a 7-point Likert-type scale and may be used individually or collectively (Duncan & McKeachie, 2005). Of the 81 items included in the 15 subscales, 8 require reverse coding. In this study, two of the Learning Strategies subscales were evaluated—the Metacognitive Self-Regulation (MSR) subscale and the Effort Regulation (ER) subscale. Pintrich et al. (1993) proposed that the MSR scale with 12 items assessed one’s aptitude for three general regulatory processes (i.e., planning, monitoring, and regulating) and the ER subscale with four items assesses learners’ ability to manage resources. More specifically, the ER subscale assesses one’s aptitude for controlling their attention and effort regardless of distractions and level of interest (Pintrich, et al., 1991, 1993). The MSR and ER scales were selected for the purposes of this study for both statistical and theory-based reasons.

Although these scales frequently appear in the research literature, there are a number of statistical issues regarding the MSR subscale and the ER subscale. First, multiple studies as well as the data for the current research do not support that the two subscales measure unique constructs. There is some debate as to whether or not metacognitive and effort regulation components of self-regulation are distinguishable (Malpass, O’Neil, & Hocevar, 1999). Although the separation makes conceptual sense, evidence suggests that respondents are unable to distinguish between the two proposed strategies through self-report (Malpass et al., 1999). For example, Pintrich and DeGroot (1990) reported that their data did not support the separation of the ER subscale and the MSR subscale into two separate scales. Furthermore, Yap (1993) found that the two subscales lacked discriminant validity.

The second issue is the latent structure of the MSLQ. Pintrich et al. (1991, 1993) conducted a confirmatory factor analysis to evaluate the latent construct assessed by the Learning Strategies Scale. They reported the goodness-of-fit index (GFI) and adjusted goodness-of-fit index (AGFI) of .78 and .75, respectively. In the MSLQ manual, they noted that the GFIs were “not stellar” but noted that “they are, nevertheless, quite reasonable values” (Pintrich et al., 1991, p. 79), deeming them acceptable enough to move forward. However, using current standards, those indices indicate that the original data fit the model poorly and the proposed latent structure was very problematic. Therefore, the purpose of this study was to investigate the latent construct of the MSR and the ER subscales. Because the original latent structure was problematic, exploratory factor analysis (EFA) was used in the current study to explore a new latent structure.

The MSR and ER scales were also selected for the purposes of this study for theoretical reasons. First, the MSR and ER scales were specifically designed to assess the multidimensions of self-regulation: motivation, cognition (strategies), and behavior (regulative action; Pintrich, 1995; Zimmerman, 1986, 1989). By assessing planning, monitoring, regulating, and resource management, MSR and ER scales assess the critical components of Zimmerman’s (1986) three-phase model of self-regulation: forethought (planning), performance control (regulating, monitoring, and resource management), and self-reflection (monitoring and planning). Finally, the MSR and ER scales were selected for this study because of statistical issues.

Academic Self-Regulation

A vast body of research indicates that high levels of academic self-regulation result in positive academic outcomes for students (e.g., Bembenutty, 2008; Cleary & Zimmerman, 2004; Pintrich & DeGroot, 1990; Zimmerman, 1990). This critical learning construct from social learning theory is a multifaceted, complex process that consists of a triad of dimensions and a three-phase cycle (Cleary & Zimmerman, 2004; Zimmerman, 1986, 1989, 1998, 2000). Generally, self-regulation refers to volitional monitoring and coordination of cognitive activities (Zimmerman, 1986). More specifically, academic self-regulation pertains not only to conscious and effortful awareness, coordination of cognitive and metacognitive thought processes, and learning strategies but also to the selection and application of appropriate strategies directed at achieving learning goals (Duncan & McKeachie, 2005).

This process reflects that students are motivationally, cognitively, and behaviorally active participants in the learning process (Bembenutty, 2008; Pressley et al., 1990; Zimmerman, 1989). The motivational components of self-regulation include perceived competence, self-efficacy, and autonomy. The cognitive aspects of self-regulation include planning, organizing, self-instructing, self-monitoring, and self-evaluating throughout the learning process. The behavioral dimension of self-regulation involves the selection, structuring, and/or creation of environments that facilitate learning (Zimmerman, 1986, 1989, 1998, 2000).

A better understanding of the aforementioned dimensions and theoretical structure of self-regulation is provided by Zimmerman’s (1998, 2000) three-phase cyclical model of self-regulation, which is rooted in social–cognitive theory and research. In accordance with this theory of self-regulation, academic self-regulation is defined as one’s thoughts, feelings, and behaviors that are planned, self-generated, and cyclically adjusted to attain one’s personal learning goals. The three phases of Zimmerman’s (1989, 2000) cyclical model of self-regulation are forethought, performance control, and self-reflection.

Forethought processes precede action and effort, setting the stage for learning (Cleary & Zimmerman, 2004). Forethought processes include goal setting and strategic processes, and are affected by the following motivational constructs: self-efficacy, goal orientation, control beliefs, and task value. In the forethought phase of self- regulation, highly self-regulated learners are those who value the task at hand, proactively set goals, determine a plan of action, and approach the task in an attentive, efficacious fashion determined to attain their aspirations for learning (Cleary & Zimmerman, 2004; Zimmerman, 2000).

The second stage of the three-phase model of self-regulation is performance control. Performance control processes of self-regulation are self-control and self-observation. These two performance control processes occur during learning efforts, affecting both attention and action. Self-control processes involve the orchestration of learning efforts or skill performance (e.g., self-instruction and attention focusing). Self-observation involves students monitoring their learning efforts and skill performance (e.g., questioning and self-assessment; Zimmerman, 1989). The performance control processes are critical to self-regulation and successful learning because the learner collects information that will be used to evaluate the effectiveness of the initial strategic plan and to advance future learning efforts (Cleary & Zimmerman, 2004; Zimmerman, 2000).

The final phase of the self-regulation cycle is self-reflection. Self-reflection processes occur after performance efforts and involve self-judgments and self-reactions. In the self-reflection phase of self-regulation, highly self-regulated learners are those who appraise their performance relative to self-oriented standards (e.g., previous paper grade), attribute less than desirable performance to faulty strategies (e.g., using too few sources for writing a paper), and make tactical modifications to strategies prior to the next learning task (e.g., use more resources for future papers). Self- reflection processes are important to future learning efforts because these processes influence one’s interpretation and response to the learning experience and influence the next self-regulatory cycle (Cleary & Zimmerman, 2004; Zimmerman, 1998, 2000).

Self-regulation is the cumulative product of this three-phase cycle and related processes. Positive academic outcomes result from one’s ability to successfully manage learning and academic performance through these self-regulatory processes (Bembenutty, 2008; Cleary & Zimmerman, 2004; Pintrich & DeGroot, 1990; Zimmerman, 1990). Fortunately for educational stakeholders and students alike, academic self- regulation is a trainable student characteristic (Boekaerts & Corno, 2005; Cleary & Zimmerman, 2004; Perels et al., 2005; Schunk, 1996; Zimmerman, 1989). Because academic self-regulation is both malleable and critical to successful learning outcomes, it is important that educational researchers and classroom educators have both a statistically and theoretically sound means of assessing students’ aptitude for academic self-regulation (Boekaerts & Corno, 2005). Therefore, the purpose of the remainder of this article is to provide both.

Although the MSR subscale and the ER subscale from the MSLQ–Learning Strategies Scale appear numerous times with acceptable reliability for each in empirical studies of self-regulation (Duncan & McKeachie, 2005), it is important to reexamine the issue of how well the items that assess the targeted latent constructs load onto the factors expected to measure the targeted latent constructs. According to the MSLQ manual (Pintrich et al., 1991), the Learning Strategies Scale has shown data–model fit poorly. Therefore, if there is information available (i.e., based on EFA results) on where the departure from the hypothesized model lies, it could provide some clues as to what a better structure might be.

Pintrich et al. (1991, 1993) used EFA procedures in their original study, but they did not provide information regarding the following issues: (a) What types of factor extraction methods were performed in conjunction with the factor analysis procedure (e.g., principal-axis factoring [PAF], maximum likelihood, unweighted least square, generalized least square, etc.)? (b) What type of rotation methods were used with the factor analysis procedure? (c) How many items were removed during the development stage and why? (d) What were the final factor loadings? Moreover, after recognizing the inadequate nature of the latent structure on the MSLQ–Learning Strategies Scale, Pintrich et al. (1993) tried to use the modification index to modify their model, but they stated, “Modifying the model did not substantively improve the overall fit indices (e.g., the GFI increased from .779 to .789; the RMR decreased from .078 to .076)” (p. 809). Therefore, the fit indices that resulted from the modification index continued to illustrate an inadequate data–model fit for the MSLQ-Learning Strategies Scale. This provides a clue that the hypothesized model has shown not only a certain degree of misspecification but also a serious problem on its latent structure. In light of the criterion and outcomes discussed, the EFA procedure is needed for investigating a better latent structure on both the MSR and ER subscales.

In addition to these statistical issues, profound sociocultural changes have occurred since the original development of this instrument in the late 1980s and early 1990s. For example, the past 20 years have borne witness to the multifaceted and ever-increasing role of technology in education, and greatly increased emphasis on both formative and summative assessment in education. By reevaluating the structure of the MSR and the ER scales using contemporary students, modified scales that are both reliable and more finely calibrated to measure the target latent construct, one’s aptitude for academic self-regulation, may emerge from existing MSLQ items.

Thus, the purpose of this article is to statistically and theoretically reexamine the MSR and the ER scales. Thus, a reevaluation of Pintrich et al.’s (1991, 1993) original analyses and an examination of new data from the MSR and ER scales are discussed. As a result of this work, the current authors propose modified, more statistically sound means of using some of the MSR and ER items to assess one’s aptitude for academic self-regulation. It is important to note that Pintrich et al. (1991, 1993) assert that the MSLQ scales and subscales may be used collectively or separately, which in conjunction with the statistical analyses used for the current study support the targeted reevaluation of the MSR and ER scales.

Method

Participants

Participants were sampled from four different mid-southern universities and were enrolled in a wide variety of courses. The first group completing the 81 items of the MSLQ consisted of 92 graduate students (13 males, 69 females) from a mid-southern university. The ethnic composition of this group was as follows: Caucasian 88.2%, African American (10.7%), and Hispanic (1.1%). Another group of students also completed the full version of the MSLQ and included 263 undergraduate students (32 males, 239 females) from three different mid-southern universities. The ethnic composition of the second group was as follows: Caucasian 59.4%, African American (31.7%), Asian (4.4%), Hispanic (1.5%), and Other (3%). All participants volunteered and ranged in age from 24 to 57 years (M = 33 years) in the first group and from 18 to 45 years (M = 24 years) in the second group.

Procedure

For the first group (n = 92), the researcher contacted participants by e-mail and asked them to take the MSLQ via an online format. For the second group (n = 263), participants took pen-and-paper versions of the MSLQ. For both groups, all responses were voluntary and anonymous.

Analysis

SAS 9.2 was used to conduct descriptive analysis and perform the EFA. Initially, there were 355 participants when two data sets merged together. However, three cases showed systematic response bias (i.e., same responses for the entire questionnaire disregarding positively or negatively worded items) and were eliminated. Additionally, two cases with missing values, which showed no evidences of violating assumptions of randomness, were also deleted. An imputation strategy was not used to deal with missing data because the missing rate was less than 1%. This was below the suggested guideline of 5% (Nosal & Nosal, 2003). Next, the total sample (N = 350) was randomly divided into two split-half samples (n = 175). The first split-half sample was used to perform an EFA and to reevaluate the psychometric properties of two latent constructs (i.e., subscales of MSR and ER). After the factor structure was revealed and appropriated items chosen with the EFA findings, Mplus 5.2 (Muthén & Muthén, 2007) was used to conduct a confirmatory factor analysis with the second split-half sample to confirm the factorial validity of the identified scale and to evaluate the hypothesized model with the GFIs according to Hu and Bentler’s (1999) recommendation.

Results

Descriptive Statistics

The means and standard deviations from the MSR and ER subscales from the first and the second split-half samples are shown in Table 1.

Table 1.

Means and Standard Deviations for Both Split-Half Samples (N = 350)

		First split-half sample (n = 175)		Second split-half sample (n = 175)
Item		M	SD	M	SD
33.	During class time I often miss important points because I’m thinking of other things.^a	3.17	1.84	3.05	1.82
36.	When reading for this course, I make up questions to help focus my reading.	4.36	2.22	4.40	2.14
37.	I often feel so lazy or bored when I study for this class that I quit before I finish what I planned to do.^a	3.40	2.01	2.98	1.82
41.	When I become confused about something I’m reading for this class, I go back and try to figure it out.	5.62	1.61	5.49	1.70
44.	If course materials are difficult to understand, I change the way I read the material.	4.70	2.03	4.68	1.89
48.	I work hard to do well in this class even if I don’t like what we are doing.	4.54	2.06	4.21	2.13
54.	Before I study new course materials thoroughly, I often skim it to see how it is organized.	4.86	1.98	5.05	1.92
55.	I ask myself questions to make sure I understand the material I have been studying in this class.	5.07	1.95	5.07	1.75
56.	I try to change the way I study in order to fit the course requirements and instructor’s teaching style.	4.06	2.03	3.68	1.99
57.	I often find that I have been reading for class but don’t know what it was all about.^a	3.67	2.21	3.89	2.22
60.	When course work is difficult, I gave up or only study the easy parts.^a	4.79	1.85	4.94	1.81
61.	I try to think through a topic and decide what I am supposed to learn from it rather than just reading it over when studying.	5.51	1.59	5.73	1.44
74.	Even when course materials are dull and understanding, I manage to keep working until I finish.	5.13	2.01	5.29	1.78
76.	When studying for this course I try to determine which concepts I don’t understand well.	4.63	2.11	5.00	1.99
78.	When I study for this class, I set goals for myself in order to direct my activities in each study period.	4.89	1.89	4.93	1.86
79.	If I get confused taking notes in class, I make sure I sort it out afterwards.^a	4.85	1.85	5.06	1.66

Reverse-scaled item.

Exploratory Factor Analysis

Several published psychometric analyses of survey instruments indicate that using positively and negatively worded items may introduce systematic measurement error that distorts analyses and the interpretation of the results (DiStefano & Motl, 2006; Horan, Distefano, & Motl, 2003; Quilty, Oakman, & Risko, 2006). Constructs conceptualized as unidimensional may appear multidimensional when positively and negatively worded items are included (i.e., positively and negatively worded items form separate factors; Motl & DiStefano, 2002; Wang, Siegal, Falck, & Carlson, 2001). For the MSLQ, Rao and Sachs (1999) conducted a confirmatory factor analysis with junior high and high school students, and indicated participants had difficulties responding properly to reverse-coded items. Therefore, based on measurement theory and preexisting research on the MSLQ, all four reverse-coded items (i.e., Items 33, 37, 57, and 60) from the MSR and ER subscales were eliminated before conducting the EFA (see Table 2).

Table 2.

Removed Reverse-Coded MSLQ Items

MSLQ Item	Statement
MSR-33	During class time I often miss important points because I’m thinking of other things.
ER-37	I often feel so lazy or bored when I study for this class that I quit before I finish what I planned to do.
MSR-57	I often find that I have been reading for class but don’t know what it was all about.
ER-60	When course work is difficult, I give up or only study the easy parts.

Note: MSLQ = Motivated Strategies for Learning Questionnaire; MSR = Metacognitive Self-Regulation; ER = Effort Regulation.

Generally, EFA requires large sample sizes. Initially, Guilford (1954) suggested a minimum sample size of 200 for recovering consistent factors. Comrey (1973) recommended researchers to obtain more than 500 participants while conducting EFA. More recently, a number of simulation studies have investigated the required minimum sample needed to yield reliable factor recovery (de Winter, Dodou, & Wieringa, 2009; Gagné & Hancock, 2006; MacCallum, Widaman, Preacher, & Hong, 2001; MacCallum, Widaman, Zhang, & Hong, 1999; Marsh, Hau, Balla, & Grayson, 1998; Velicer & Fava, 1998). Although the suggested minimum sample size varies from one study to another, the overall theoretical framework supports that to yield reliable results one must only acquire a minimum sample size of 60, a considerably reduced requirement (MacCallum et al., 1999; MacCallum et al., 2001). Accordingly, this study’s first data set of 175 participants is sufficient to perform EFA corresponding with two potential factors on 12 items.

It is important to select a proper factor extraction method under the EFA framework. This decision is particularly important because potential problems may exist due to violations of the assumption of multivariate normality. However, several researchers posit that a PAF is the most appropriate extraction technique even when this issue arises (Costello & Osborne, 2005; Fabrigar, Wegener, MacCallum & Strahan, 1999; Floyd & Widaman, 1995). Following the above recommendation, a PAF analysis with oblique rotation was performed. Given that the two self-regulation factors might be correlated to each other, the oblique (i.e., promax) rotation, which allows factors to correlate, was chosen over the orthogonal (i.e., varimax) rotation (Floyd & Widaman, 1995; Thompson, 2004).

For the purposes of investigating and reevaluating the latent self-regulation constructs, the number of factors was not initially constrained (i.e., not set to 2). Using the recommendation of Reise, Waller, and Comrey (2000), both a scree test (Cattell, 1966) and a parallel analysis (Horn, 1965) were performed in order to determine the number of factors to extract. With the first split-half sample, both the parallel analysis and the scree test indicated a clear two-factor solution. Factor loadings and communalities are shown in Table 3.

Table 3.

Exploratory Factor Analysis Two-Factor Solution with the First Split-Half Sample

Item	Factor 1	Factor 2
36	.72	−.14
44	.61	.02
74	.59	.13
48	.56	−.01
55	.49	.10
56	.37	−.30
78	.31	.27
76	−.13	.71
41	−.08	.52
79	.17	.50
61	.30	.38
54	.30	.33

Although results indicated a two-factor solution, a comparison of the current latent structure and Pintrich et al.’s (1991, 1993) original model reveals that the current results do not align with the original model. Several items displayed inconsistent factor loadings across the original and current results. As described earlier, the original model did not fit the data well in the Pintrich et al. (1991, 1993) study. Therefore, the current authors propose two shorter, modified academic self-regulation scales with a stronger data–model fit.

According to Comrey and Lee (1992), as well as Floyd and Widaman (1995), factor loadings that exceed .40 are generally considered meaningful. Therefore, items that did not load substantially on any factor (i.e., a factor loading of less than .40) were deleted from the scale. This criterion resulted in further elimination of Items 54, 56, 61, and 78 from the self-regulation subscales. Next, the EFA was performed again on the remaining eight items. Table 4 presents the factor loadings and communalities for two latent factors identified as General Strategies for Learning (GSL) and Clarification Strategies for Learning (CSL). The GSL consisted of five items that included questions related to general self-regulation strategies. The second set of items was labeled as CSL and consisted of three items that included questions related to identifying and clarifying confusion and misunderstandings in the learning process. Cronbach’s alpha was calculated for these shorter, modified self-regulation subscales. An alpha of .60 or greater was deemed acceptable based on the literature that suggests for a group-level assessment of psychological constructs, particularly those in the initial phases of investigation, an alpha of .60 indicates acceptable levels of internal consistency (e.g., DeVellis, 1991; Hays, Anderson, & Revicki, 1998; Nunnally, 1967; Weiner, Freedheim, Graham, & Naglieri, 2003). The reliability coefficient for the overall scale was .73 and the coefficients for the GSL and CSL subscales were .74 and .61, respectively.

Table 4.

Exploratory Factor Analysis Two-Factor Solution on the Shortened Version of Self-Regulation Scale With the First Split-Half Sample

Item	Factor 1	Factor 2
36	.65	−.18
44	.65	.002
74	.61	.10
48	.59	−.01
55	.47	.12
76	−.09	.61
41	−.02	.52
79	.21	.52

New Self-Regulated Learning Scales

Two new scales emerged based on the results of the EFA. The first scale that emerged from the existing MSR and ER scales was the GSL scale (see Table 5). The second scale that emerged from the existing MSR and ER scales was the CSL scale (see Table 6).

Table 5.

General Strategies for Learning (GSL)

GSL item	Statement	Original scale and number
1	When reading for this course, I make up questions to help me focus my reading.	MSR-36
2	If course materials are difficult to understand, I change the way I read the material.	MSR-44
3	I work hard to do well in this class even if I don’t like what we are doing.	ER-48
4	I ask myself questions to make sure I understand the material I have been studying in this class.	MSR-55
5	Even when course materials are dull and uninteresting, I manage to keep working until I finish.	ER-74

Note. MSR = Metacognitive Self-Regulation; ER = Effort Regulation.

Table 6.

Clarification Strategies for Learning (CSL)

CSL item	Statement	Original scale and number
1	When I become confused about something I’m reading for this class, I go back and try to figure it out.	MSR-41
2	When studying for this course I try to determine which concepts I don’t understand well.	MSR-76
3	If I get confused taking notes in class, I make sure I sort it out afterwards.	MSR-79

Note. MSR = Metacognitive Self-Regulation.

Confirmatory Factor Analysis

A confirmatory measurement model was constructed using the two-factor structure with the second split-half sample. Assessments of model fit were based on multiple criteria including tests and interpretations of individual parameters as well as overall model fit indices. As suggested by Hu and Bentler (1999), the following criteria were used to indicate good data–model fit: >.95 for CFI, <.08 for standardized root mean square residual (SRMR), and <.06 for root mean square error of approximation (RMSEA). The results indicated that the current authors’ initially hypothesized two-factor model fit the data well, χ²(19) = 32.54, p = .027; GFI = .96; CFI = .93; SRMR = .062; RMSEA = .051, 90% confidence interval (CI) = .000-.089. In fact, MacCallum, Roznowski, and Necowitz (1992) cautioned that “when an initial model fits well, it is probably unwise to modify it to achieve even better fit because modifications may simply be fitting small idiosyncratic characteristics of the sample” (p. 501). Therefore, this two-factor model was psychometrically acceptable and no further modifications were required. Examination of the completely standardized solution for this hypothesized model is shown in Figure 1. In summary, the proposed two-factor model is supported well by our data.

Figure 1.

Completely standardized solution for Motivated Strategies for Learning Questionnaire self-regulation model

Discussion

Two modified scales resulted from these research efforts—the GSL scale and the CSL scale. These modified scales theoretically align with the latent constructs Pintrich et al. (1991, 1993) originally sought to assess through the use of the MSR and ER scales: planning, monitoring, regulating, and resource management. Furthermore, the two modified scales address the statistical issues associated with the MSR and the ER scales. Therefore, the statistical reevaluation of the MSR scale and the ER scale resulted in a reconfiguration of original MSLQ items into two modified scales that assess the latent constructs originally targeted by Pintrich et al. (1991, 1993). The following discussion presents a theoretical rationale for the removal of four original MSLQ items, an explanation of the GSL, and an explanation of the CSL. Additionally, a discussion ensues supporting that the GSL scale and the CSL scale assess the same underlying self-regulatory characteristics as the MSR scale and the ER scale.

Deleted MSLQ Items

Four items (54, 56, 61, and 78) were removed from the original MSLQ items because of ambiguous loading on the GSL scale and the CSL scale (see Table 7). The items neither loaded strongly on the two new factors nor on a forced third factor; therefore, one cannot distinguish among them clearly. An examination of the items’ verbiage revealed that each of the items contained both general strategy and clarification strategy construct-related material. For example, Item 61 contained a general strategy—thinking through new material to create a plan, and a clarification strategy—clarifying what is critical to learn. The general strategies presented in the remaining three items were as follows: Item 54—preplanning and skimming, Item 56—adapting to new demands, and Item 78—goal setting. The clarification strategies in these remaining items were as follows: Item 54—clarifying structure of new material, Item 56—clarifying teacher and course demands, and Item 78—clarifying the purpose of studying.

Table 7.

Deleted Motivated Strategies for Learning Questionnaire (MSLQ) Items

MSLQ item	Statement
54	Before I study new course materials thoroughly, I often skim it to see how it is organized.
56	I try to change the way I study in order to fit the course requirements and instructor’s teaching style.
61	I try to think through a topic and decide what I am supposed to learn from it rather than just reading it over when studying.
78	When I study for this class, I set goals for myself in order to direct my activities in each study period.

Alternatively, the combination of the two themes of general and clarification strategies could be viewed as merging into a third construct that theoretically overlapped with the constructs measured by the GSL, the CSL, and the original MSLQ MSR, ER, and Organization scales. The construct of organization manifested in the deleted items in the following ways: Item 54—identifying the organization of new material, Item 56—organizing study behavior to fit teaching style or course demands, Item 61—organizing key concepts from new material, and Item 78—organizing goals and activities. However, these items did not load significantly on a forced third factor. It is as if the deleted items contained three separate yet related constructs of general strategies, clarification strategies, and organization strategies. Therefore, it is hypothesized that the conceptual dissonance as well as the interrelatedness of the constructs represented in these items, led to ambiguous factor loadings. This theoretical ambiguity accompanied by statistical ambiguity resulted in the removal of the items.

Modified Self-Regulated Learning Scales

Two modified scales emerged from the statistical analyses preformed in this study. Both scales assess the same regulatory abilities as the original MSR and ER scales: planning, monitoring, regulating, and resource management. However, the GSL scale does so at a more global level, whereas the CSL scale does so at a more specific level.

General Strategies for Learning scale

The first modified scale that emerged from the existing MSR and ER scales is the GSL scale (see Table 8). The GSL assesses learners’ aptitude for engaging in academic self-regulation, using the processes of planning, monitoring, regulating, and resource management. In this case, academic self-regulation is best defined as a form of learning that is guided by metacognition, partially intrinsically motivated, and strategic in nature (Winne, 1997; Zimmerman, 1990). Thus, this scale indicates one’s aptitude for metacognitively and strategically managing resources and self-regulating learning in a manner that requires intrinsic motivation.

Table 8.

General Strategies for Learning (GSL) Scale

GSL item	Statement	Original scale and number
1	When reading for this course, I make up questions to help me focus my reading.	MSR-36
2	If course materials are difficult to understand, I change the way I read the material.	MSR-44
3	I work hard to do well in this class even if I don’t like what we are doing.	ER-48
4	I ask myself questions to make sure I understand the material I have been studying in this class.	MSR-55
5	Even when course materials are dull and uninteresting, I manage to keep working until I finish.	ER-74

Note. MSR = Metacognitive Self-Regulation; ER = Effort Regulation.

Clarification Strategies for Learning scale

The CSL scale was the second modified scale to emerge from the existing MSR and ER scales (see Table 9). The CSL assesses learners’ aptitude for monitoring learning progress and identifying misunderstandings during the learning process, creating a strategic plan, managing resources such as attention and effort, and engaging in strategies to clarify direction of learning as well as areas of confusion. Misunderstandings of material interfere or detract from the learning process (Alexander & Murphy, 1998). Because teachers may not always be able to identify areas in which students are confused or hold misunderstandings, the ability of students to identify these implicit issues in their own black box is critical. This scale indicates one’s aptitude for identifying areas of misunderstanding and confusion and one’s aptitude for engaging in strategies to ameliorate any such impediments to learning. Thus, the CSL scale measures the academic self-regulatory processes assessed by the original MSR and ER scales (planning, monitoring, regulating, and resource management) but measures these processes as they specifically pertain to clarification strategies (the GSL scale is a more global measure of academic self-regulation).

Table 9.

Clarification Strategies for Learning (CSL) Scale

CSL item	Statement	Original scale and number
1	When I become confused about something I’m reading for this class, I go back and try to figure it out.	MSR-41
2	When studying for this course I try to determine which concepts I don’t understand well.	MSR-76
3	If I get confused taking notes in class, I make sure I sort it out afterwards.	MSR-79

Note. MSR = Metacognitive Self-Regulation; ER = Effort Regulation.

Conceptual Overlap of Original and Modified Scales

Not only do the modified GSL and CSL scales measure the same facets of academic self-regulation measured by the original MSR and ER scales, but they also do so in a more statistically sound manner while still assessing the same underlying self-regulatory processes. Pintrich et al. (1991, 1993) assert that the MSR scale assessed one’s aptitude for planning, monitoring of comprehension, and regulating, and that the ER scale assessed one’s aptitude for resource management. However, the current authors propose that the MSR items assessed not only one’s aptitude for planning, monitoring, and regulating but also one’s aptitude for resource management. For example, creating questions to focus reading (MSR Item 36) reflects planning, monitoring, and resource management (i.e., managing attention and effort).

Furthermore, effort regulation, such as when one works to completion on a task regardless of interest level (ER Item 74), also reflects regulation (i.e., adjusting attention similar to how one might adjust reading speed depending on the task at hand). This overlap in underlying meaning provides a theoretical basis for the reorganization of the items into the proposed GSL and CSL scales. The GSL scale and the CSL scale assess one’s aptitude for planning, monitoring, regulating, and managing resources similar to the original MSLQ scales. However, the GSL scale reflects a more global measure of one’s aptitude for self-regulation, and the CSL scale assesses one’s aptitude for self-regulation as it pertains to identifying and clarifying misunderstandings in the learning process.

Limitations and Future Research

This study was limited by the small number of male respondents (12%). However, the current literature pertaining to gender differences in self-regulation is both mixed and primarily focused on junior high and high school students. For example, Malpass et al. (1999) found no significant differences in self-regulation between male and female 10th through 12th grade students. Hong, O’Neil, and Feldon (2005) found that only 3% of the variance in the state form of self-regulation was explained by gender “indicating no substantially important differences” (p. 280). With regard to the trait form of self-regulation, they found no significant differences between the sexes. However, Ablard and Lipschultz’s (1998) qualitative investigation of self-regulation gender differences in high-achieving students supported gender differences. Zimmerman and Martinez-Pons (1990) also reported significant gender differences in mathematically gifted students’ use of self-regulatory strategies in 5th-, 8th-, and 11th-grade students. A review of research on gender differences in postsecondary students’ self-regulatory abilities also revealed mixed results. For example, Ruban and McCoach (2005) found no significant gender differences in postsecondary self-regulation, whereas Rasheed (2005) found significant gender differences in self-regulation among college students with learning disabilities.

In light of the mixed findings regarding gender differences in self-regulation and the small number of males in the current study (12%), further research with regard to the proposed scales and gender differences is recommended. In particular, it will be important to validate this measure and explore gender differences in both high-performing students and learning-disabled students. Also, future research efforts should further explore the GSL and CSL scales across different populations. For example, future research could explore the validity of the scales with high school students.

Conclusions

According to researchers, self-regulation powerfully influences academic success (e.g., Bembenutty, 2008; Cleary & Zimmerman, 2004; Pintrich & DeGroot, 1990; Zimmerman, 1990). Pintrich et al.’s (1991, 1993) work undoubtedly contributed to educator’s understanding of this critical component of successful learning. However, after two decades of social change and statistical advancements, it is important to revisit the MSLQ.

The product of this research drafted on the monumental work of Pintrich et al. (1991, 1993). The purpose of this study was not to replace but rather to revamp the existing MSLQ scales designed to assess metacognitive self-regulation and effort regulation. The GSL and CSL scales assess the same components of self-regulation measured by the original MSR and ER scales: planning, monitoring, regulating, and resource management. The GSL does so at a more global level, and the CSL does so at a more specific level. The GSL and CSL scales provide a revised, theoretically and statistically sound means of measuring students’ aptitude for the pivotal academic ability of self-regulation. Such a tool may prove useful to educators and researchers alike as they attempt to better understand and address students’ academic self-regulation.

Footnotes

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

The author(s) received no financial support for the research, authorship, and/or publication of this article.

References

Ablard

K. E.

Lipschultz

R. E.

(1998). Self-regulated learning in high-achieving students: Relations to advanced reasoning, achievement goals, and gender. Journal of Educational Psychology, 90, 94-101.

Alexander

P. A.

Murphy

P. K.

(1998). The research base for APA’s learner-centered psychological principles. In Lambert

N. M.

McCombs

B. L.

(Eds.), How students learn: Reforming schools through learner-centered education (pp. 25-60). Washington, DC: American Psychological Association.

Barkley

R. A.

(2010). Differential diagnosis of adults with ADHD: The role of executive function and self-regulation. Journal of Clinical Psychiatry, 71, e17. Retrieved from http://www.psychiatrist.com/abstracts/abstracts.asp?abstract=201007/0710e17.htm

Bembenutty

(2008). Self-regulation of learning and test anxiety. Psychology Journal, 5, 122-139.

Boekaerts

Corno

(2005). Self-regulation in the classroom: A perspective on assessment and intervention. Applied Psychology, 54, 199-231.

Cattell

R. B.

(1966). The scree test for the number of factors. Multivariate Behavioral Research, 1, 245-276.

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, 537-550.

Comrey

A. L.

(1973). A first course in factor analysis. New York, NY: Academic Press.

Comrey

A. L.

Lee

H. B.

(1992). A first course in factor analysis (2nd ed.). Hillsdale, NJ: Lawrence Erlbaum.

10.

Costello

A. B.

Osborne

J. W.

(2005, July). Best practices in exploratory factor analysis: Four recommendations for getting the most from your analysis. Practical Assessment, Research & Evaluation, 10(7), 173-178. Retrieved from http://pareonline.net/getvn.asp?v=10&n=7

11.

DeVellis

R. F.

(1991). Scale development: Theory and applications. Newbury Park, NJ: Sage.

12.

de Winter

J. C. F.

Dodou

Wieringa

P. A.

(2009). Exploratory factor analysis with small sample sizes. Multivariate Behavioral Research, 44, 147-181.

13.

DiStefano

Motl

R. W.

(2006). Further investigating method effects associated with negatively worded items on self-report surveys. Structural Equation Modeling, 13, 440-464.

14.

Duncan

T. G.

McKeachie

W. J.

(2005). The making of the motivated strategies for learning questionnaire. Educational Psychologist, 40, 117-128.

15.

Fabrigar

L. R.

Wegener

D. T.

MacCallum

R. C.

Strahan

E. J.

(1999). Evaluating the use of exploratory factor analysis in psychological research. Psychological Methods, 4, 272-299.

16.

Floyd

F. J.

Widaman

K. F.

(1995). Factor analysis in the development and refinement of clinical assessment instruments. Psychological Assessment, 7, 286-299.

17.

Gagné

Hancock

G. R.

(2006). Measurement model quality, sample size, and solution propriety in confirmation factor models. Multivariate Behavioral Research, 41, 65-83.

18.

Greene

J. A.

Azevedo

(2007). Adolescents’ use of self-regulatory processes and their relation to qualitative mental model shifts while using hypermedia. Journal of Educational Computing Research, 36, 125-148.

19.

Guilford

J. P.

(1954). Psychometric methods (2nd ed.). New York, NY: McGraw-Hill.

20.

Hays

Anderson

Revicki

D. A.

(1998). Assessing reliability and validity of measurement in clinical trials. In Staquet

M. J.

Hays

R. D.

Fayers

P. M.

(Eds.), Quality of life assessment in clinical trials (pp. 162-189). New York, NY: Oxford University Press.

21.

Heatherton

T. F.

Wagner

D. D.

(2011). Cognitive neuroscience of self-regulation failure. Trends in Cognitive Sciences, 15, 132-139.

22.

Hong

O’Neil

H. F.

Feldon

(2005). Gender effects on mathematics achievement: Mediating role of state and trait self-regulation and test anxiety. In Gallagher

A. M.

Kaufman

J. C.

(Eds.), Gender differences in mathematics: An integrative psychological approach (pp. 264-293). Cambridge, England: Cambridge University Press.

23.

Horan

P. M.

DiStefano

Motl

R. W.

(2003). Wording effects in self-esteem scales: Methodological artifact or response style? Structural Equation Modeling, 10, 435-455.

24.

Horn

J. L.

(1965). A rationale and test for the number of factors in factor analysis. Psychometrika, 30, 179-185.

25.

Bentler

(1999). Cutoff criterion for fit indexes in covariance structure analysis: Conventional criteria versus new alternatives. Structural Equation Modeling, 6, 1-55.

26.

MacCallum

R. C.

Roznowski

Necowitz

L. B.

(1992). Model modifications in covariance structure analysis: The problem of capitalization on chance. Psychological Bulletin, 111, 490-504.

27.

MacCallum

R. C.

Widaman

K. F.

Preacher

K. J.

Hong

(2001). Sample size in factor analysis: The role of model error. Multivariate Behavioral Research, 36, 611-637.

28.

MacCallum

R. C.

Widaman

K. F.

Zhang

Hong

(1999). Sample size in factor analysis. Psychological Methods, 4, 84-99.

29.

Malpass

J. R.

O’Neil

H. F.

Hocevar

(1999). Self-regulation, goal orientation, self-efficacy, worry, and high-stakes math achievement for mathematically gifted high school students. Roeper Review, 21, 281-289.

30.

Marsh

H. W.

Hau

K. T.

Balla

J. R.

Grayson

(1998). Is more ever too much? The number of indicators per factor in confirmatory factor analysis. Multivariate Behavioral Research, 33, 181-220.

31.

Motl

R. W.

DiStefano

(2002). Longitudinal invariance of self-esteem and method effects associated with negatively worded items. Structural Equation Modeling, 9, 562-578.

32.

Muthén

L. K.

Muthén

B. O.

(2007). Mplus user’s guide (3rd ed.). Los Angeles, CA: Muthén & Muthén.

33.

Nunnally

J. C.

(1967). Psychometric theory. New York, NY: McGraw-Hill.

34.

Nosal

(2003, June). Three-parameter Weibull generator for replacing missing observations. Paper presented at the meeting of the International Conference on Statistics and Related Fields, Honolulu, HI.

35.

Paris

S. G.

Winograd

(2001). The role of self-regulated learning in contextual teaching: Principles and practices for teacher preparation. Washington, DC: U.S. Department of Education. Retrieved from http://www.ciera.org/library/archive/2001-04/0104parwin.htm

36.

Perels

Dignath

Schmitz

(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, 17-31.

37.

Perels

Gurtler

Schmitz

(2005). Training of self-regulatory and problem-solving competence. Learning and Instruction, 15, 123-139.

38.

Pintrich

P. R.

(1995). Understanding self-regulated learning. In Pintrich

P. R.

(Ed.), Understanding self-regulated learning (pp. 3-12). San Francisco, CA: Jossey-Bass.

39.

Pintrich

P. R.

DeGroot

E. V.

(1990). Motivational and self-regulated learning components of classroom academic performance. Journal of Educational Psychology, 82, 33-40.

40.

Pintrich

P. R.

Smith

D. A. F.

Garcia

McKeachie

W. J.

(1991). A manual for the use of the Motivated Strategies for Learning Questionnaire (MSLQ). Ann Arbor: National Center for Research to Improve Postsecondary Teaching and Learning, The University of Michigan.

41.

Pintrich

P. R.

Smith

D. A. F.

Garcia

McKeachie

W. J.

(1993). Reliability and predictive validity of the motivated strategies for learning questionnaire (MSLQ). Educational and Psychological Measurement, 53, 801-813.

42.

Pressley

Woloshyn

Lysynchuk

L. M.

Martin

Wood

Willoughby

(1990). A primer of research on cognitive strategy instruction: The important issues and how to address them. Educational Psychology Review, 2, 1-58.

43.

Quilty

L. C.

Oakman

J. M.

Risko

(2006). Correlates of the Rosenberg Self-Esteem Scale method effects. Structural Equation Modeling, 13, 99-117.

44.

Rao

Sachs

(1999). Confirmatory factor analysis of the Chinese version of the motivated strategies for learning questionnaire. Educational and Psychological Measurement, 59, 1016-1029.

45.

Rasheed

S. A.

(2005). Self-determined skills in postsecondary students with learning disabilities and their correlates (Doctoral dissertation). Available from ProQuest Dissertations and Theses database. (Publication No. 3195338)

46.

Reise

S. P.

Waller

N. G.

Comrey

A. L.

(2000). Factor analysis and scale revision. Psychological Assessment, 12, 287-297.

47.

Ruban

L. M.

McCoach

D. B.

(2005). Gender differences in explaining grades using structural equation modeling. Review of Higher Education, 28, 475-502.

48.

Schunk

D. H.

(1996). Goal and self-evaluative influences during children’s cognitive skill learning. American Educational Research Journal, 33, 359-382.

49.

Schunk

D. H.

Zimmerman

B. J.

(1998). Self-regulated learning: From teaching to self-reflective practice. New York, NY: Guilford Press.

50.

Stoeger

Ziegler

(2008). Evaluation of a classroom based training to improve self-regulation in time management tasks during homework activities with fourth graders. Metacognition and Learning, 3, 207-230.

51.

Thompson

(2004). Exploratory and confirmatory factor analysis: Understanding concepts and applications. Washington, DC: American Psychological Association.

52.

Velicer

W. F.

Fava

J. L.

(1998). Effects of variable and subject sampling on factor pattern recovery. Psychological Methods, 3, 231-251.

53.

Wang

Siegal

H. A.

Falck

R. S.

Carlson

R. G.

(2001). Factorial structure of Rosenberg’s self-esteem scale among crack-cocaine drug users. Structural Equation Modeling, 8, 275-286.

54.

Weiner

I. B.

Freedheim

D. K.

Graham

J. R.

Naglieri

J. A.

(Eds.). (2003). Handbook of psychology: Assessment psychology. Hoboken, NJ: Wiley.

55.

Winne

P. H.

(1997). Experimenting to bootstrap self-regulated learning. Journal of Educational Psychology, 89, 1-14.

56.

Winne

P. H.

Perry

N. E.

(2000). Measuring self-regulated learning. In Boekaerts

Pintrich

P. R.

Zeidner

(Eds.), Handbook of self-regulation (pp. 531-566). San Diego, CA: Academic Press.

57.

Yap

E. G.

(1993). A structure model of self-regulated learning in math achievement (Unpublished doctoral dissertation). University of Southern California, Los Angeles.

58.

Zimmerman

B. J.

(1986). Becoming a self-regulated learner: Which are the key subprocesses? Contemporary Educational Psychology, 11, 307-313.

59.

Zimmerman

B. J.

(1989). A social cognitive view of self-regulated academic learning. Journal of Educational Psychology, 81, 329-339.

60.

Zimmerman

B. J.

(1990). Self-regulated learning and academic achievement: An overview. Educational Psychologist, 25, 3-18.

61.

Zimmerman

B. J.

(1998). Academic studying and the development of personal skill: A self-regulatory perspective. Educational Psychologist, 33, 73-86.

62.

Zimmerman

B. J.

(2000). Attaining self-regulation: A social cognitive perspective. In Boekaerts

Pintrich

P. R.

Zeidner

(Eds.), Handbook of self-regulation (pp. 13-39). San Diego, CA: Academic Press.

63.

Zimmerman

B. J.

Martinez-Pons

(1990). Student differences in self-regulated learning: Related grade, sex, and giftedness to self-efficacy and strategy use. Journal of Educational Psychology, 82, 51-59.