Using pictures, they asked the children to identify the medications they had taken while they had Cries syndrome. Their classmates without Cries syndrome were also asked to use the pictures to identify medications taken during the same time period. The investigators then asked each child to classify each medication taken as good-tasting or bad-tasting medication. The data on spankings were obtained from the primary caregiver. The investigators found the following data:

	Cries Syndrome	Controls
Percentage of children who reported taking bad-tasting medication	90%	10%
Average number of spankings per day	1	2
Average number of television viewing hours per day	8 (range 5-12)	2 (range 0 to 4)

Among the 20 variables, analyzed one at a time, the above variables were the only ones that were statistically significant using the usual statistical methods. The P values were 0.05 except for the hours of television viewing, which had a P value of 0.001.