Capturing multiple interaction effects in L1 and L2 object-naming reaction times in healthy bilinguals: a mixed-effects multiple regression analysis

Table 6 Analysis 2 (German only): final model fixed effects

Terms	Estimate	Std. Error	T-obt	95% CI lower	95% CI upper	K&R df	p-value	Sign.
Intercept	1.218E−03	2.459E−04	4.950	8.269E−04	1.788E−03	8.4	9.77E−04	***
Run Num (run 2)	9.064E−05	7.519E−06	12.053	7.603E−05	1.055E−04	1329	< 2e−16	***
trial_number	− 6.108E−07	1.091E−07	− 5.598	− 8.212E−07	− 3.916E−07	1376	2.61E+08	***
GermanRun1%Correct	− 3.631−04	2.836E−04	−1.280	−9.200E−04	1.901E−04	8.1	0.2359
wordChoice (others)	− 2.916E−05	1.083E−05	− 2.688	− 5.039E−05	−8.429E−06	1434	0.0073	**
Gender	−1.068E−04	4.645E−05	− 2.298	− 1.970E−04	− 1.540E−05	5.9	0.0624
log10WF	1.548E−04	5.266E−05	2.938	5.285E−05	2.563E−04	1402	0.0034	**
firstPhonemeDiff	− 5.199E−05	1.208E−05	− 4.281	− 7.596E−05	− 2.777E−05	1410	1.98E−05	***
gender*firstPhonemeDiff	4.277E−05	1.853E−05	2.302	6.279E−06	7.984E−05	1388	0.0215	*
GermanRun1%Correct*log10WF	− 1.295E−04	5.904E−05	− 2.193	−2.437E−04	−1.494E−05	1346	0.0285	*

This table provides an overview reflecting the final statistical model used for the within-German reaction time (RT) comparisons. The degrees of freedom to determine the p-values were calculated using Kenward and Roger’s method. The p-value indicates that the gender factor was only marginally significant. However, the 10,000 bootstrap CI that did not include 0 suggests that the gender factor was reliable

ISSN: 1471-2202