Key points are not available for this paper at this time.
One‐minute transcripts of 30 university students' first in‐class public speeches earlier (Mulac males higher on Dynamism. For the present study, these transcripts were analyzed linguistically by 11 trained coders for 35 language features selected as potential discriminators of speaker gender. Discriminant analysis results showed that a combination of 20 of these features could account for 99% of the between‐gender variance, permitting 100% accuracy of gender prediction. Multiple regression analyses demonstrated that 13 of the gender‐discriminating language features predicted the three attributional dimensions in ways consistent with the Gender‐Linked Language Effect.
Mulac et al. (Sun,) studied this question.