(Post last updated 28 December 2024)
Review panel summary
The Metacognitive Awareness Inventory (MAI) consists of 52 items on a continuous response scale and was developed to assess students' metacognitive awareness across eight dimensions of metacognition: declarative knowledge, procedural knowledge, conditional knowledge, planning, information management strategies, monitoring, debugging strategies, and evaluation.
Evidence based on internal structure validity was demonstrated through confirmatory factor analysis of a two-factor structure—knowledge of cognition and regulation of cognition. Several studies have provided validity evidence of the MAI's data based on its relations with other variables, including task difficulty, self-confidence, familiarity [2], exposure to simulations [4], GPA (4), and metacognitive skills [3]. Reliability of the MAI in the form of single administration has been supported across studies, with reasonable coefficient alpha estimates for each subscale. The instrument is available with operational definitions for each construct [1], and a modified version has been explored [3].
Recommendations for use
The use of a two-subscale model of the MAI instrument is well-supported, while evidence for the finer structure (i.e., the 8 dimensions) remains limited. Therefore, given the current evidence, it is suggested that only scores for ‘knowledge of cognition’ and ‘regulation of cognition’ be calculated and reported. Further research is needed to provide evidence to support response process validity, particularly given potential language shifts since the instrument's original development in 1994.
Evidence supports the consequential validity of MAI data, as scores improved following a targeted metacognitive training intervention. However, the instrument is now 30 years old, and all prior studies have used paper-and-pencil data collection. With the increased prevalence of online data collection, examining the scale’s performance in an online format is recommended.
The instrument’s applicability with diverse populations is supported, as previous studies indicate its suitability across various educational settings.
Details from panel review
The initial version of the instrument was pilot tested with 120 items. The final 52 items were selected based on the pilot test results. Items with extreme mean scores and items with large score variability were dropped. The final item selection was done so that each of the proposed six subscales contained four items. The final version of the instrument was administered and the data were analyzed with unrestricted exploratory factor analysis using orthogonal and oblique solutions. The proposed 8 factor structure was not supported. Some factors were aligned with the knowledge of cognition category, while the other factors aligned with the regulation of knowledge category. A forced two-factor solution revealed support for the two major categories. Evidence for this two-factor structure was found in a second experiment, with different subjects [1]. No evidence related to response process validity was reported.
Single administration reliability data in the form of coefficient alpha suggests that the two factor scales are reliable [1]. Validity evidence in the form of relation to other variables is provided by the scores on the two scales being highly, positively, correlated with other measures related to metacognition [1-3]. Additionally, evidence based on the consequence of testing comes in the form of the change in pre/post scores in a study where the treatment was metacognitive training [3].
While the MAI was originally developed with introductory psychology students, it was used in a variety of educational settings, including general chemistry [2], upper division psychology [3], and engineering [4]. However, no statistical evidence of measurement invariance has been reported.
References
[1] Schraw G. and Dennison, R.S. (1994). Assessing Metacognitive Awareness. Contem. Educ. Psych., 19, 460-475.
[2] Gulacar, O. and Bowman, C.R. (2014). Determining what our students need most: exploring student perceptions and comparing difficulty ratings of students and faculty. Chem. Educ. Res. & Pract. 15, 587-593.
[3] Terlecki, M.S. & McMahon, A. (2018). A Call for Metacognitive Intervention: Improvements Due to Curricular Programming in Leadership. J. Lead. in Educ. 17(4), 130-145.
[4] Elliott, L.J., Aqlan, F., Zhao, R., & Janney, M.S. (2020). Assessment of Metacognitive Skills in Design and Manufacturing. Amer. Soc. for Eng. Eudc. Ann. Conf. Proc., Paper 31121.