Investigating stable and dynamic aspects of student motivation using generalizability theory

Abstract

Background The Academic Motivation Scale (AMS) is a commonly used 28‐item measure of motivation orientation. However, the scale has not been examined using methods to distinguish between state and trait aspects of motivation. We applied Generalizability Theory to appraise the psychometric properties of the AMS and to differentiate items measuring state or trait motivation within the AMS. Methods One hundred and thirty medical students completed the 28‐item AMS at three time points. Generalizability Theory was applied to examine the generalizability of the AMS (total and sub‐scales) and to estimate state and trait aspects of student motivation. Results The overall AMS showed high generalizability (G = .93) in assessing student motivation across student populations and occasions, whereas the AMS subscales demonstrated low levels of generalizability. The majority of the AMS items scored high for the state component index suggesting movement between motivational priorities while maintaining the overall stability of the individual motivation levels. Conclusions The total AMS is suitable as a reliable measurement of motivation among student populations with high generalizability of scores across students and assessment occasions. Measuring trait motivation using the total AMS score instead of subscale scores reduces measurement error and optimises instrument reliability.

Keywords

• academic motivation scale
• assessment
• generalizability theory
• reliability
• state and trait
• student motivation