Time Use During First Year of College Predicts Participation in High-Impact Activities During Later Years

The value of a college education depends on the experiences and opportunities students engage in over the course of their college career (Kuh, Hu, & Vesper, 2000; Kuh, Kinzie, Buckley, Bridges, & Hayek, 2006), and the transition to college is a particularly important time as it sets the stage for students' academic and personal success in the years to come (Tinto, 2007). To increase student engagement, many universities are adopting high-impact educational practices that include study abroad opportunities, faculty mentoring, internships, service learning, challenging coursework, and research experiences; these institutions are also intentionally promoting high-impact cocurricular activities such as community service and leadership positions in campus organizations (Kuh, 2008). How students spend their time during their first year may have implications for engagement in high-impact activities during their third and fourth years. For instance, many first-year students engage in high rates of drinking, socializing, and passive entertainment such as watching movies (Padilla-Walker, Nelson, Carroll, & Jensen, 2010; Schulenberg & Maggs, 2002). Evidence suggests that these activities are associated with students' engagement in high-impact activities such as volunteering (Finlay, Ram, Maggs, & Caldwell, 2011) during that same year; however, few studies have examined first-year students' time use and their engagement in high-impact activities during their third and fourth years. As colleges and universities increase their investment in high-impact activities, understanding which first-year experiences predict higher levels of participation later could prove useful for early intervention. With this longitudinal study, we explored how students' time use (e.g., volunteering, napping, going to bars and parties) during their first year of college predicted participation in a subset of high-impact activities (civic engagement, study abroad, leadership) and course selection (easy or difficult) in their third and fourth years of college.

METHOD

Participants

Participants were part of a longitudinal study of undergraduate students at a large Northeastern university (Patrick, Maggs, & Lefkowitz, 2015). Eligibility requirements included being a full-time first-year student under the age of 21, being a US citizen or permanent resident, and residing within 25 miles of campus. The study used a longitudinal burst design. Students responded to a baseline survey and 14 consecutive daily surveys for 7 consecutive semesters starting in their first semester. Students completed all surveys online. Students were selected using stratified sampling to recruit a diverse sample with respect to gender and the four largest race/ethnicity categories. A total of 744 participants provided consent and completed the Semester 1 (S1) baseline survey for a response rate of 65.6%.

In our analysis we used data regarding students' time use from the daily surveys at Semester 1 (S1) and Semester 2 (S2), and data regarding students' participation in high-impact activities from the semester survey at Semesters 5, 6, and 7 (S5, S6, and S7). Students completed S1 and S2 during their first year at the university, completed S5 and S6 during their third year, and completed S6 during the first semester of their fourth year. The retention rate (i.e., percentage completing a variable of interest on the S5, S6, or S7 surveys) was 88.8% (n = 661). Due to missing data (i.e., study attrition and items that participants skipped), the final analytic sample (n = 652) was 87.6% of the full sample and 53.5% female. Students could identify as more than one race or ethnicity; thus, the sample was 45.6% White / European American, 29.1% Asian American / Hawaiian / Pacific Islander, 26.2% Hispanic / Latino American, 21.0% Black / African American, and 2.6% Native American / American Indian. We used nine t tests and six chi-square tests to determine whether participants in the analytic sample differed from participants not in the analytic sample on S1 variables. Participants in the analytic sample were more likely to be female, χ²(1, n = 744) = 15.62, p < .001, and tended to spend less time napping in S1, t(724) = –4.13, p < .001, and playing video games in S1, t(724) = –2.00, p < .05, than participants not in the analytic...