Table 3.

Prediction of number of drinks in a week (Model 1)

Predictors	B (SE)	t	p
Intercept of alcohol use	1.44 (0.29)	4.90	.000
Biweekly-level (L1) predictors of alcohol use
Prior assessment negative evaluationa	−0.10 (0.03)	−3.32	.000
Negative evaluations x Class year	0.13 (0.07)	1.86	.063
Prior assessment positive evaluationa	0.08 (0.05)	1.72	.086
Positive evaluations x Class year	0.03 (0.10)	0.28	.782
Prior assessment number of negative consequencesa	−0.04 (0.02)	−1.75	.082
Prior assessment number of positive consequencesa	0.01 (0.01)	1.16	.249
Prior assessment sum of drinksa	0.02 (0.004)	5.46	.000
Week of semester	−0.06 (0.01)	−5.93	.000
School year week (vs. break)	0.24 (0.15)	1.60	.110
Class year	−0.02 (0.05)	−0.46	.644
Person-level (L2) predictors of alcohol use (intercept effects)
Gender	−0.65 (0.09)	−6.91	.000
Race/Ethnicity	0.42 (0.10)	4.30	.000
School	0.14 (0.07)	1.98	.048
Cohort	0.06 (0.06)	1.11	.267
Average negative evaluation across weeks	−0.27 (0.09)	−2.91	.004
Average positive evaluation across weeks	0.52 (0.11)	4.85	.000
Average number of positive consequences across weeks	−0.02 (0.03)	−0.72	.475
Average number of negative consequences across weeks	0.63 (0.07)	8.75	.000
Person-level (L2) moderators of evaluation effects (slope effects)
Gender x (L1) negative evaluations	0.06 (0.06)	1.01	.314
Gender x (L1) positive evaluations	0.12 (0.10)	1.23	.219

Note. Gender (0.4= female, −0.6 = male), Race/Ethnicity (0 = Non-White, 1 = Non-Hispanic White); three School sites coded 1–3; three Cohorts coded 1–3; School Year Week (0 = Break Week, 1 = School Year Week); Class Year (−0.5 = First-Year, 0.5 = Sophomore); removal of cross-level interaction effects of gender (L2) on evaluations (L1) did not alter the pattern or significance of model findings

^aPerson-centered variable that reflects deviations from the participant’s average score on that variable across weeks.

Bold values are significant at p < .05.