Examining the construct validity of academic self -regulation using the Survey of Academic Self -regulation (SASR)
This purpose of this study was to examine the construct validity of Academic Self-Regulation (ASR) using a new self-report questionnaire—the Survey of Academic Self-Regulation (SASR). Several ASR construct-validity issues have endured in the empirical literature, many due to the use of ASR self-report measures with dubious psychometric qualities; the SASR was developed in an attempt to overcome these construct-validity issues.
The SASR was first pilot tested on a convenience sample of college students (N = 205) from upstate New York to assess its initial reliability and validity. It was then administered to a larger sample (N = 491) from the same institutions to further establish its reliability and validity, and to examine the enduring ASR construct validity issues. Correlational, factor, multiple regression, and reliability analyses were conducted in both studies.
A six-factor structure for the SASR was cross-validated across the two studies, and these factors correlated well with those from two popular ASR self-report measures—the Learning and Study Strategies Inventory (LASSI) and Motivated Strategies Learning Questionnaire (MSLQ)—providing evidence of convergent (and discriminant) validity. Most SASR factors were significant predictors of achievement (grades and GPA), and total SASR score was related in simple (linear) and complex (nonlinear) ways to assorted contextual and individual characteristic variables. These findings were mostly consistent with previous ASR research using self-report measures, although the more complex findings were unique to this study.
Implications for this research include the following: the moderate number of factors found for the SASR (6), LASSI (7), and MSLQ (8) indicate a need to revise existing theories that postulate more ASR components (application to theory); the complex findings from this study need cross-validation, as does the SASR factor structure, on more diverse samples (application to research); and lastly, the SASR thus far appears be a more reliable and valid self-report measure of ASR than either the LASSI or the MSLQ, and as such, may prove to be a more authentic and efficient measure for practitioners to use for the identification, diagnosis, promotion, and/or remediation of ASR skills (application to practice).