Table 2.

Estimates for change in outcome measures due to tray removal

	Control dining hall		Intervention dining hall		Diff-in-diff estimate
	β	p value	β	p value	β	p value
No. of lunch servings	0.088	0.086	−0.100	0.031	−0.188	0.006
No. of drink servings	0.108	0.014	−0.299	<.0001	−0.407	<.0001
No. of salad servings	−0.048	0.017	0.021	0.309	0.070	0.015
No. of dessert servings	0.003	0.757	−0.024	0.039	−0.027	0.076
No. of dishes with leftover	−0.021	0.561	−0.176	<.0001	−0.153	0.001
No. of entrée with leftover	−0.008	0.811	−0.162	<.0001	−0.152	0.005

The first four columns of the tables are the estimated results for the control (first two column) and intervention (second two column) dining hall regressed separately. The estimates shown are for the binary post intervention variable. The difference in difference estimate on the relative change between the control and intervention dining hall is shown in the last two columns