Table 10.

Logistic mixed-effect regression models for (ING) in Dataset C (516 tokens) (Model III repeated from Table 8).

	Model III: human coders (N = 516 tokens)			Model V: radial SVM predictions (N = 516 tokens)
	Est.	Std. Err.	p	Est.	Std. Err.	p
(Intercept)	5.86	1.40	< 0.0001***	3.47	0.75	< 0.0001***
Corpus (DCB, vs. DCA)	0.25	0.61	0.6846	−0.69	0.32	0.0334*
Gender (male, vs.female)	1.17	0.55	0.0326*	0.66	0.27	0.0155*
AgeGrp (AG1, vs. AG4)	−0.10	0.82	0.9024	−1.11	0.44	0.0126*
AgeGrp (AG2, vs. AG4)	−1.38	0.86	0.1058	−1.30	0.43	0.0026**
AgeGrp (AG3, vs. AG4)	−1.16	0.78	0.1358	−0.44	0.37	0.2362
SEC (SE2, vs. SE1)	−1.09	0.64	0.0888.	0.01	0.32	0.9704
SEC (SE3, vs. SE1)	−3.67	0.77	< 0.0001***	−0.83	0.34	0.0140*
GramCat (N-like, vs. V-like)	−0.93	0.49	0.0573.	−0.34	0.31	0.2643
Word Len (# Sylls)	−1.39	0.42	0.0010***	−0.75	0.25	0.0022**

^***p < 0.001,

^**p < 0.01,

^*p < 0.05, .p < 0.1.