Board 136: Preliminary Validity Evidence for a Brief Measure of Engineering Identity

• Published in: Association for Engineering Education - Engineering Library Division Papers
• Main Authors: Major, Debra A, Katelyn R. Reynoldson M.S., Hu, Xiaoxiao, Burleson, Seterra D, Shryock, Kristi J
• Format: Conference Proceeding
• Language: English
• Published: Atlanta American Society for Engineering Education-ASEE 23.06.2018
• Subjects: Colleges & universities; Convergence; Correlation analysis; Engineering; Engineering education; Factor analysis; Graduates; Student retention; Students; Validity

Considering national initiatives to increase the overall number of engineering graduates, improving the persistence of students to remain in engineering disciplines through to graduation has become a pivotal strategy. Although lack of interest or ability has been used to explain why some leave engineering prior to graduating, the degree to which engineering becomes central to a student’s self-concept (i.e., engineering identity) has been supported as a better explanatory factor in retention-related outcomes. The focus of this paper is the development of a concise quantitative instrument that measures undergraduate students’ engineering identity, thus allowing for further understanding of engineering retention through identity. Survey responses from participants entering and throughout their first year of an undergraduate engineering program at a large southeastern university were used to develop and validate the 5-item measure. In developing the measure, eleven identity items either adapted from the career identity literature or generated for this project were evaluated. First-semester survey responses to these items were analyzed to determine which items were most suitable to include in a 5-item measure of engineering identity. An exploratory factor analysis indicated that the identity items represented a single construct, a finding that was confirmed with confirmatory factor analyses (CFA). The identity measure was reduced to five items by using CFA modification indices, examining factor loadings, and emphasizing items from the existing literature. Finally, two other survey samples were used to validate the 5-item measure, including a sample of participants entering the engineering program and a sample of second-semester engineering students. Fit indices for the 5-item measure were satisfactory for all samples, and Cronbach’s alpha for the identity measure was above recommended levels for all three samples. Convergent validity was demonstrated by significant positive correlations between the 5-item identity measure and each of the three dimensions of embeddedness (fit, links, and sacrifice). Discriminant validity was demonstrated through nonsignificant correlations between identity and participant SAT scores. Lastly, predictive validity was demonstrated through significant relationships between identity and major satisfaction one semester later. Predictive validity was further demonstrated with retention data; identity measured upon entry and after one semester in the engineering program was significantly negatively related to first-year engineering retention. Discriminant, convergent, and predictive validity for the 5-item measure will continue to be assessed as longitudinal data are gathered regarding student engineering identity in their second year in the engineering program. The paper describes progress to date in the validation of this measure, future steps, and the potential utility of the measure. Having a concise, validated measure of identity will be valuable for quick assessment of student engineering identity and gaining further understanding of retention in engineering disciplines. This work was supported by a National Science Foundation IUSE grant.