The effects of food advertising and cognitive load on food choices

BMC Public Health

Table 3 Effect of food advertising on the number of calories chosen

		All calories		Healthy calories		Unhealthy calories
	N	Coeff*	*p-value*	Coeff*	*p-value*	Coeff*	*p-value*
	N	[95% CI]	*p-value*	[95% CI]	*p-value*	[95% CI]	*p-value*
Full sample	351	65 [10 – 121]	*0.02*	15 [–34 – 65]	*0.55*	70 [19 – 122]	*0.01*
Stratification by experimentally manipulated cognitive load
Low-cognitive-load sub-sample	173	36 [–46 – 118]	0.39	15 [–57 – 88]	0.68	34 [–38 – 105]	0.36
High-cognitive-load sub-sample	178	94 [19 – 169]	*0.01*	15 [–53 – 83]	0.67	107 [33 – 181]	*0.01*
Sub-stratifications by parental SES
Low cognitive load
High SES	68	69 [–75 – 213]	0.34	58 [–70 – 185]	0.37	38 [–78 – 154]	0.51
Low SES	76	-45 [–153 – 63]	0.41	-52 [–147 – 43]	0.28	16 [–99 – 132]	0.78
High cognitive load
High SES	78	34 [–92 – 160]	0.59	28 [–79 – 134]	0.61	27 [–105 – 160]	0.68
Low SES	72	90 [–22 – 203]	0.11	-51 [–163 – 60]	0.36	143 [37 – 249]	*0.01*

*The coefficient (Coeff) is the difference in total calories chosen by those in the food advertising group from the total chosen by those in the non-food-advertising group. [95% Confidence Intervals (CI) in brackets].
Note: For each outcome, the results of 7 separate Tobit regressions are reported. In each regression the intervention status is the only variable. Significant results in bold.

ISSN: 1471-2458