Table 3. Exploratory effects.

Each section of the table reports general linear hypothesis tests of ratings on the two foil dimensions for two target song types. Comparisons are between a target song type (e.g., healing songs for “to mourn the dead”) and the other three song types and are listed in descending order of effect size. Bolded results are statistically significant at alpha = .05. The statistics in this table correspond to the visualizations in the left-hand panels of Figure 3.

	Mdiff	95% CI	F(1,114)	p	z-score
Healing songs as used “to mourn the dead”
vs. dance songs	0.73	[0.34, 1.13]	13.8	.000320	0.93
vs. lullabies	0.38	[−0.01, 0.77]	3.68	.0576	0.48
vs. love songs	0.29	[−0.10, 0.68]	2.11	.149	0.36
Love songs as used “to tell a story”
vs. lullabies	0.33	[0.11, 0.54]	8.79	.00368	0.74
vs. healing songs	0.26	[0.04, 0.49]	5.57	.0199	0.60
vs. dance songs	0.19	[−0.03, 0.41]	2.91	.0910	0.43