Table 2.

Percentage of schools that used strategies to promote school meals in 2013–2014, with different percentages shown by state law status, where significantly different in a multiple regression model to predict use of each strategy^a

Variable	Overall		% by State Lawd		P value	Total unweighted n schools
	%	Unweighted n schools	In states with no law	In states with a law
Collected suggestions from students about the school nutrition services program	53.6	242	49.8	67.1	0.017	411
Collected suggestions from students’ families about the school nutrition services program	26.7	109			0.56	409
Conducted taste tests with students	48.3	217	44.1	63.5	0.016	413
Conducted taste tests with students’ families	11.2	41	9.5	18.7	0.046	411
School has committee with students who provide suggestions for the school nutrition services program	20.4	95			0.38	400
Made menus available to studentsb	97.3	403				414
Made information available to students on the nutrition and caloric content of foods available to them	68.1	285			0.08	410
Placed posters or other materials promoting healthy eating habits on display in cafeteria	94.1	385			0.09	414
Placed posters or other materials promoting healthy eating habits on display in school	48.4	186			0.76	401
Included nutrition services topics during school announcements	53.0	215	48.1	70.4	0.002	405
Included articles about the school nutrition services program in a school newsletter, newspaper, website, or other publication	65.7	274			0.24	406
Made menus available to families	96.4	395			0.17	413
Made information available to families on the nutrition and caloric content of foods available to students	68.9	278			0.279	409
Made information available to families on the school nutrition services program	83.8	345			0.11	410
Met with a parents’ organization, such as the PTA,c to discuss the school nutrition services program	26.8	103			0.92	402
Invited family members to a school meal	56.5	224	53.1	70.7	0.015	411

^aP values for each outcome are based on significance test for the state law variable in a series of 16 multivariate logistic regressions, where each outcome is regressed on all demographic covariates (school level, size, locale, region, student race/ ethnicity, and socioeconomic composition) plus state law status.

^bThis model could not be computed to assess whether prevalence was associated with state law, due to the very high prevalence (limited variability) of this outcome.

^cPTA=parent teacher association.

^dAdjusted percentage is the percentage of schools using each practice, after accounting for school demographic characteristics (school level, size, locale, region, student race/ethnicity, and socioeconomic composition). Percentages in states with or without laws are shown only where the difference is statistically significant at P<0.05 or better.