Table 7.

J(apanese), N(orwegian), E(nglish), and I(talian); S(peech) and N(on-)S(peech). The top values in each cell were obtained from a model in which all the listeners’ data are pooled, while the bottom values are obtained from jackknifing the data (see the text for explanation). Italics mark significance for the top values; significance can be inferred from the 95% confidence intervals supplied for the jackknife estimates. Values listed are the differences in log likelihood ratio (Δ-2LLR) between models in which silence and vowel duration are independent variables and those which include the interaction between them, and β values for the vowel by silence duration interaction and the independent variables representing silence duration and vowel duration.

Lg.	Stimuli	Δ-2LLR	Vowel by Silence	Silence	Vowel
J	S	0.828	−.00014	.1321	.0562
		−1.595 ± 8.222	−.00015 ± .00060	.1414 ± .0414	.0761 ± .0496
	NS	2.639	.000242	.1075	.0344
		3.0276 ± 9.2839	.000225 ± .000457	.1075 ± .0293	.0355 ± .0412
N	S	0.03	.00002	.0174	−.0346
		−0.844 ± .942	.00003 ± .00021	.0160 ± .0258	−.0343 ± .0441
	NS	2.379	.00015	.0289	.0063
		3.896 ± 6.707	.00017 ± .00024	.0269 ± .0334	.0043 ± .0211
E	S	1.519	−.00009	.0375	−.0306
		1.172 ± 6.127	−.00011 ± .00022	.0389 ± .0201	.0298 ± .0273
	NS	1.688	−.0001	.0351	−.0243
		1.150 ± 6.359	−.00016 ± .00020	.0307 ± .0253	.0214 ± .0269
I	S	4.036	−.00031	.1010	−.0053
		6.631 ± 8.355	−.00031 ± .00037	.1010 ± .0394	−.0053 ± .0291
	NS	3.111	−.00020	.0747	.0460
		3.761 ± 11.197	−.00022 ± .00045	.0751 ± .0376	.0478 ± .0338