Table 3.

Summary of the fixed effects in two mixed logit models (with and without memory as a covariate)

Model	Fit	Predictor	Coef.	SE	Wald Z	p > |Z|
Not controlling for vSTM M2	−2LL = 2439
	χ2(11) = 110.15	Intercept	−.454	.246	−1.843	.065
	p < 2.2e−16	Embedding = RB	.708	.311	2.275	.023*
		Gap = Sub	1.146	.316	3.624	<.001***
		Group = Adults	2.453	.318	7.705	<.001***
		Group = DLD	−.656	.282	−2.324	.020*
		RB : Sub	−1.699	.443	−3.835	<.001***
		RB : Adults	−.849	.385	−2.209	.027*
		RB : DLD	.204	.323	.633	.527
		Sub : Adults	−1.265	.389	−3.250	.001**
		Sub : DLD	−.406	.331	−1.227	.220
		RB : Sub : Adults	1.256	.530	2.371	.018*
		RB : Sub : DLD	.813	.454	1.788	.074
Controlling for vSTM M3	−2LL = 2435
	χ2(15) = 114.98	Intercept	−.454	.246	−1.845	.065
	p < 2.2e−16	Embedding = RB	.707	.311	2.271	.023*
		Gap = Sub	1.147	.317	3.621	<.001***
		Group = Adults	1.977	.431	4.583	<.001***
		Group = DLD	−.352	.340	−1.036	.300
		vSTM	.056	.035	1.583	.114
		RB : Sub	−1.698	.443	−3.830	<.001**
		RB : Adults	−.172	.518	−.331	.743
		RB : DLD	−.227	.391	−.579	.562
		Sub : Adults	−.757	.523	−1.447	.148
		Sub : DLD	−.730	.399	−1.831	.067
		RB : vSTM	−.079	.041	−1.938	.053
		Sub : vSTM	−.060	.041	−1.451	.147
		RB : Sub : Adults	.728	.717	1.015	.310
		RB : Sub : DLD	1.144	.548	2.087	.037*
		RB : Sub : vSTM	.061	.055	1.108	.268

Note. ‘:’ indicates an interaction term.

^*- p < .05,

^**- p < .01,

^***- p < .0001.

−2LL - −2 log-likelihood (deviance). RB - right-branching, Sub - subject gap, vSTM – verbal short-term memory (Digit Span), DLD – Developmental Language Disorder. Fit indices are presented relative to the null model with just random effects of items and subjects on the intercept.