This study examines children's gender schema regarding language styles in compliments using oral discourse completion tasks. Twenty-five Mandarin-speaking children (aged 9-12) provided compliments in normal and imitated (opposite-gender) styles. Machine learning analyzed lexical, discourse-pragmatic, and discourse-semantic features. Results show discourse-pragmatic features are more influenced by gender ideologies than lexical features. Girls' imitated compliments showed significantly less positivity. Findings support gender-differentiated language styles in pre-adolescents, particularly girls' awareness of boys' less prosocial language in compliments.

Xinyu Liao, Yanhui Zhang