Mixed-effects linear regression testing the fixed effects of presence of variable contexts (reference level = <i>absent</i>), turn duration, mean word length and speech rate and the random intercept of speaker on length of speech planning time per turn among young South Midland females in the Switchboard Corpus.

Speech planning time log transformed. Linear predictors centered and scaled. Treatment contrast coding. Model fit by maximum likelihood. Probability calculated using Satterthwaite approximation for degrees of freedom. Coefficients reported in log-odds. Variation inflation factors <1.23.