Table 1. Binary logistic GEE results for the Path ROI (per time window of interest).

	By-participants (N = 43)							By-items (N = 36)
Window	Effect	Est.	SE		p	ΔQIC	Est.	SE		p	ΔQIC
Verb	Speed	+.306	.178	2.67	.10		+.307	.187	2.64	.11
	Time	+.058	.008	15.18	.001	+37.5	+.062	.010	24.11	.001	−48.3
	S×T	−.017	.010	2.60	.11		−.016	.011	2.41	.12
Path	Speed	−.500	.132	10.33	.001		−.501	.116	12.48	.001
	Time	−.002	.007	0.04	.85	−108.7	−.002	.005	0.09	.76	−200.0
	S×T	+.005	.009	0.36	.55		+.006	.006	1.01	.31
Goal	Speed	−.498	.166	5.92	.02		−.502	.146	8.87	.003
	Time	−.029	.014	4.61	.04	−113.2	−.030	.012	5.15	.03	−193.4
	S×T	+.004	.019	0.05	.83		+.005	.012	0.15	.70
Post-sent.	Speed	−.364	.189	3.35	.07		−.362	.142	5.67	.02
	Time	+.004	.009	1.56	.21	−58.6	+.004	.002	4.23	.04	−183.5
	S×T	+.013	.012	1.28	.26		+.013	.008	2.26	.13

Table shows GEE modelling results for occurrences of looks to the Path ROI as a function of Verb Speed (fast vs. slow) and Time (continuous predictor referring to the 50 ms time-bins per window). Shown are the by-participant/by-item GEE parameter estimates and corresponding SEs (both in logit units) along with generalised score chi square statistics and related p-values (at df = 1). As for interactions (S×T), a positive parameter estimate indicates a more positive slope for the Time predictor in the fast than in the slow Verb Speed condition. ΔQIC refers to the goodness of fit of the Verb Speed×Time model in relation to a competitor Agent-Verb Suitability×Lexical Frequency×Time model (negative ΔQICs indicate superior fit of the Verb Speed×Time model).