The Discriminative Lexicon: A Unified Computational Model for the Lexicon and Lexical Processing in Comprehension and Production Grounded Not in (De)Composition but in Linear Discriminative Learning

Table 1

Linear mixed model fit to paired associate learning scores with the correlation of the row vectors in of the paired words as predictor. Treatment coding was used for factors. For the youngest age group, is not predictive, but all other age groups show increasingly large slopes compared to the slope of the youngest age group. The range of is ; hence larger values for the coefficients of and its interactions imply worse performance on the PAL task.

A. parametric coefficients	Estimate	Std. Error	t-value	p-value
intercept	3.6220	0.6822	5.3096	0.0001
	0.3058	0.1672	1.8293	0.0691
age=39	0.2297	0.1869	1.2292	0.2207
age=49	0.4665	0.1869	2.4964	0.0135
age=59	0.6078	0.1869	3.2528	0.0014
age=69	0.8029	0.1869	4.2970	0.0001
sex=male	-0.1074	0.0230	-4.6638	0.0001
:age=39	0.1167	0.0458	2.5490	0.0117
:age=49	0.2090	0.0458	4.5640	0.0001
:age=59	0.2463	0.0458	5.3787	0.0001
:age=69	0.3239	0.0458	7.0735	0.0001
B. smooth terms	edf	Ref.df	F-value	p-value
random intercepts word pair	17.8607	18.0000	128.2283	0.0001