Effects of talker gender and face masks on the speech recognition of six-year-old children in a classroom

Miji KwonWonyoung Yang^*

• Gwangju University, Gwangju, Republic of Korea

Although mandatory wearing of face masks for three years owing to COVID-19 might have strongly affected children's language development, its effects on their speech recognition based on the talker's gender remain unknown. This study examined how face mask usage affects children's speech recognition, focusing on the interaction between the talker's gender and the child listener's characteristics under realistic acoustic conditions with room reverberation and background noise. Speech recognition was assessed in 43 six-year-old children who had worn masks for two or more years during preschool. Auralisation techniques using male and female professional voice actors' recordings under varying room reverberation and background noise conditions were used for the assessment. The assessment revealed significant talker gender effects, both with and without face masks. Gender interactions were observed, with girls demonstrating significant differences in speech recognition scores based on talker gender, whereas boys showed no such variations. Face masks attenuated the talker gender effect on speech recognition. Listener gender showed no significant impact in the overall analysis; however, thicker face masks were associated with improved speech recognition at lower reverberation times and noise levels. Reverberation time significantly affected speech recognition only in younger children (mean age: 74 months). Face masks reduced vowel working space areas across both genders. Thus, optimising the acoustic environment is crucial for younger children wearing face masks in educational settings. This study has important implications for classroom acoustics and educational spaces during periods of mandatory mask usage.