Table 2.

Multilevel Models of Change in Emotional Support From University Friends Over Time, With Between-Person Effects of Key Predictors on Time 1 Emotional Support

Predictor	γ	SE
Intercept	8.55***	.49
Gender	.60*	.24
Age	−.08*	.03
Marital status	−.47	.30
Part time versus full time	1.87***	.33
Military status	−1.10*	.52
Time	.37*	.16

Note. Convergence criterion = .001. Gender (male = −1, female = 1), marital status (unmarried = −1, married = 1), and full- versus part-time student status (−1 = part time, 1 = full time) were effect coded. Military status (0 = civilian, 1 = student service member/veteran) was dummy coded. Between-person variables (Level 2), such as gender, age, marital status, part-time versus full-time status, and military status, estimate effects on level of the outcome, which was centered to Time 1. Within-person variables (Level 1), such as time, estimate the rate of change in the outcome.

^*p < .05.

^***p < .001.