Measuring Civic Engagement in First-Year College Students: A Confirmatory Factor Analysis

Background and Purpose:

College mental-health professionals serve students at a time when depression and substance abuse are increasing among undergraduate populations (Swaner, 2007). Binge drinking and illicit use of prescription drugs is prevalent (Crandford, et al., 2009; McCabe, et al, 2006). College-students also report high levels of depression and anxiety (Silverman, 2009). Civic engagement may serve to protect against these behaviors and promote well-being (Astin & Sax, 1998; Michelsen et al., 2002). However, current measures of civic engagement tend to be limited in scope. Most scales developed over the last 15 years focus on attitudes and intentions and lack items that measure actions. Recent research does suggest that increasing percentages of students anticipateengaging in civic activity (Hurtado & DeAngelo, 2012). However, anticipating or intending to partake in civic activities does not mandate that they occur. Therefore, the development of parsimonious measures of civic engagement that measure both attitudes and behaviors is needed in order to further research on civic engagement’s efficacy for promoting well-being. By combining measures of civic-oriented activities and civic attitudes the scale produced in this study aims to offer an instrument that provides a comprehensive assessment of civic engagement. The purpose of this study was to investigate the factor structure for measures of civic mindedness and civic-oriented activities in a sample of first-year college students.

Methods:

A cohort of first-year, undergraduate students at a private university in the U.S. Mountain West region were invited to take an online survey administered during Fall and Spring terms.  The survey included a civic mindedness scale (Eyler & Giles, 1999) and a civic-oriented activities scale (BTtoP Project Team, n.d.). Item responses consisted of a 5-point Likert scale. In SPSS 22, data from the Spring survey (n=226) were used to investigate the factor structure of the measures with exploratory factor analysis (EFA) using principal axis factoring and Promax rotation.  Second, construct validity of the factor structure from the EFA was assessed with confirmatory factor analysis (CFA) and weighted-least squares means and variance adjustment (WLSMV) estimation in Mplus (7.3) using data from the Fall survey (n=352).

Results:

EFA found a single factor for the civic mindedness scale (5 items, alpha=.698) and the civic-oriented activities scale (8 items, alpha=.737).  CFA results showed model-to-data fit for the civic mindedness scale (chi-square(4)=2.072, p=.7225; RMSEA=.000; CFI=1.000; TLI=1.000) and the civic-oriented activities scale (chi-square(16)=32.888, p=.0076; RMSEA=.055; CFI=.979; TLI=.963). The measurement model with the two latent factors of civic mindedness and civic-oriented activities correlated, also achieved model fit (chi-square(59)=94.817, p=.0021; RMSEA=.042; CFI=.976; TLI=.968).   

Conclusions and Implications:

The EFA and CFA confirmed that both the civic mindedness and civic acts subscales combined to measure the latent construct of civic engagement. Many previous civically oriented scales were limited in their scope by focusing solely on civic attitudes and intentions. The development of a parsimonious reliable scale that encompasses both civic attitudes and actions provides researchers with an improved tool for measuring civic engagement.