Table 3. Continuous coarticulatory model vs. categorical (front/non-front) model, steady-state F2 and F1

DV	Model	AIC	BIC	χ²(1) for added category	p-value
F2std	Continuous only (F2on, F1on, Dur1)	489.6	500.0	—	—
F2std	+ Frontness category	490.6	502.8	0.99	.319
F1std	Continuous only (F2on, F1on, Dur1)	497.2	507.6	—	—
F1std	+ Frontness category	499.1	511.2	0.09	.770

All models include random intercepts for speaker and a variance component for vowel. Lower AIC/BIC indicates better fit; neither comparison supports retaining the categorical term once continuous coarticulatory predictors are included.