Table 7. Speculation on Cause of Word Aversion.

Experiment	Aversion	Sound	Connotation	K	χ2(3)	χ2(1)
1 (N = 400)	Averse (n = 82)	40.2%	39.0%	.803	43.462	26.646
	Non-averse	11.3%	52.8%		V = .571	V = .315
2 (N = 377)	Averse (n = 57)	38.6%	49.1%	.710	16.694	13.555
	Non-averse	15.6%	66.9%		V = .498	V = .231
3 (N = 572)	Averse (n = 108)	51.9%	37.0%	.831	54.175	40.553
	Non-averse	19.0%	60.3%		V = .533	V = .296
4 (N = 370)	Averse (n = 49)	51.0%	38.8%	.778	39.501	28.465
	Non-averse	15.0%	66.4%		V = .566	V = .305
5 (N = 688)	Averse (n = 139)	47.5%	43.9%	.841	47.743	31.207
	Non-averse	19.9%	58.7%		V = .415	V = .293

Speculation on the cause of moist aversion in Experiments 1–5 grouped by self-reported moist-aversion. The “sound” column reflects the proportion of participants who identified the sound alone or the sound and connotation as aversive; the “connotation” column reflects the proportion of participants who identified the connotation alone as aversive. K refers to Cohen’s Kappa, a measure of inter-rater reliability. Two chi-square tests of independence are reported for each experiment: one in which all four categories of responses were taken into account, and one in which two categories were taken into account (sound, which collapsed over sound and connotation, and connotation alone; the “other” category was excluded for this second test). All ps < .001. Cramer’s V is reported as a measure of effect size.