Table 1.

Results of the statistical models

		LMM predicting tabooness/offensiveness			GLMM predicting taboo word status
Predictor type	Predictor	b	t	p	b	z	p
Intercept	Intercept	0.398	32.43	< .001	1.834	3.43	< .001
Main effects	Dummy	0.349	20.15	< .001
	Valence	−0.515	−50.77	< .001	−9.586	−19.08	< .001
	Arousal	0.496	40.68	< .001	13.111	19.65	< .001
	Concreteness	0.068	6.58	< .001	−1.322	−2.77	< .001
	AoA	0.181	20.04	< .001	−2.364	−5.81	< .001
	Corpus freq.	−0.009	−10.53	< .001	−0.309	−7.90	< .001
Interactions	Dummy: Valence	−0.209	−14.38	< .001
	Dummy: Arousal	−0.142	−8.26	< .001
	Dummy: Concreteness	−0.139	−9.55	< .001
	Dummy: AoA	−0.200	−15.65	< .001
	Dummy: Corpus freq.	−0.003	−2.39	.017
LMM predicting tabooness/offensiveness: predictors of tabooness and offensiveness ratings across samples for the dataset of all words (words produced in Study 1 and fillers). “Dummy” is a dummy variable coding for tabooness ratings (coded as 0, the reference condition) or offensiveness ratings (coded as 1); therefore, the intercept and main effects except “dummy” describe tabooness ratings, while the “dummy” effect and all interactions describe how offensiveness ratings differ from tabooness ratings. GLMM predicting taboo word status: predictors of taboo word status across labs (1: taboo, 0: filler)