Abstract:

The purpose of this study was to evaluate the psychometric properties of the Materials Concept Inventory (MCI) and then recommend revisions. The MCI was administered to a sample of 303 undergraduate engineering students enrolled in a materials engineering course taught by an instructor unrelated to this study. The MCI demonstrated adequate reliability (Cronbach's alpha was .73) and strong discriminatory power (Ferguson's delta was 0.96). A pretest MCI was given only two days into the course and was able to predict the final course grade (r = .30, p < .001). Convergent validity was also established by significantly correlating a post-test version of the MCI with the final course grade (r = .50, p < .001). Nineteen of the thirty items on the MCI met the criterion for index of item congruency, demonstrating somewhat mixed reviews by content experts. A confirmatory factor analysis conducted on the initial MCI structure failed to support a six-factor model. A revision was made to the structure of the MCI, whereby an exploratory and subsequent confirmatory factor analysis yielded a refined six-factor structure that did provide a strong fit by the empirical data [chi-squared (194) = 196.03, p = .426; and root mean square error approximation (RMSEA) was .007; and comparative fit index (CFI) was .996]. The Cronbach's alpha of the refined-MCI was 0.75, suggesting 'good' reliability. The six factors could only be interpreted by an integration of cognitive processes with content. Those six factors were labeled: (1) compare and contrast ductile and brittle material, (2) analyze structure and how it changes due to defects arising from stretching, (3) understand and analyze the structure and properties of glass, (4) analyze structure and how it changes due to defects arising from bending, (5) understand and analyze the structure and properties of metals, and (6) synthesize and apply graphs and proportions with relation to various materials engineering concepts. Results related to the psychometrics of the refined version of the Materials Concept Inventory are promising, but based on this study's results, further revision may be required.