Skills and Attitudes for College Success: A Multigroup Confirmatory Factor Analysis With First-Generation College Students

• In Event: Applying Advanced Models in Noncognitive Assessment

Abstract

The relationship between socioeconomic status (SES) and outcomes in higher education has been thoroughly demonstrated. (Bowen, Chingos, & McPherson, 2009; Terenzini, Cabrera, & Bernal, 2001). Social capital theory (Bourdieu, 1985; Coleman, 1988) is commonly used to frame impact of SES on college success. Although financial resources play a role (St. John & Tuttle, 2004; Wei & Horn, 2009), many authors (e.g., Entwistle & Astone, 2004; Perna & Titus, 2005; Walpole, 2003) have discussed that, in addition to economic capital, students “inherit” other factors relevant to college success. One of these, cultural capital (often referred to as “college knowledge”), refers to expectations of educational attainment, knowledge of how to navigate the college system, and cultural norms that are tied to college success.

However, many of the studies that explore the impact of cultural capital have done so from educational and sociological perspectives, often using qualitative approaches. The current paper applies several key constructs from psychological research to explain gaps in the preparedness and success of first-generation students (those who do not have a parent with a college degree) and non-first generation students. The first-generation demarcation is a common measure of SES in higher education, and has been related to student success (Chen, 2005).

Conscientiousness (Poropat, 1995), time management, and social support (Robbins et al., 2004) have all been meta-analytically related to student persistence. We hypothesized that low-SES students, based on differences in cultural capital, should differ in the way they conceptualize social support and institutional commitment, but not conscientiousness and time management. This differential conceptualization, we posit, specifically arises from students’ knowledge about and willingness to engage with faculty on campus (Pike & Kuh, 2005).

In order to test this hypothesis, we conducted multi-group confirmatory factor analyses (MG-CFA) on each of the constructs mentioned. First, configural invarance was supported, indicating that the overall factor structure fit reasonably for both groups. However, tests of metric invariance indicated that, to some extent, the opposite of our hypothesis was true. Metric invariance was supported for time management and social support, but was not supported for conscientiousness. This suggests that students, in fact, conceptualize attitudes toward faculty similarly, but differ in their conceptualization of the personality trait most strongly related to academic success. (Additional discussions of manifest and latent mean difference, partial metric invariance, and other analyses will be included in the full paper.)

These findings have potentially broad and significant implications. First, to our knowledge, this is the first study to examine the invariance of personality across socioeconomic student groups, and the suggestion that these students differ is a noteworthy finding. Second, these results point to the types of interventions that might facilitate success in first-generation students. Certainly, as the conference theme suggests, “to know is not enough,” and this research will hopefully, ultimately, lead to strategies and programs that can facilitate success for students that face socioeconomic disadvantages.

Authors

• Ross Edward Markle, Educational Testing Service

• Jonas Bertling, Educational Testing Service

• Priya Kannan, Educational Testing Service

• Diane Cardenas Elliott, Educational Testing Service