Identification of stuttering in bilingual Lebanese children across two presentation modes

The goals of this study were to investigate whether Lebanese speech-language pathologists (SLPs) are accurate at identifying stuttering in bilingual children, to examine whether the use of video-recordings instead of audio-recordings allows for better analyses, and to explore factors that may affect the SLPs' judgments. In phase 1, 32 SLPs listened to narrative samples in Lebanese Arabic of 6 children who do not stutter (CWNS) and 2 who stutter (CWS). They were instructed to label each child as stuttering or not, and to explain what motivated their decisions. Afterwards, they were asked to provide background information by means of a questionnaire. In phase 2, they were asked to judge the same speech samples on the basis of video-recordings, and to explain for each child which speech characteristics they relied on to make their decisions. The results showed that misidentification happens frequently, is significantly more likely to happen with CWNS than with CWS, but also varies within these categories. Moreover, the use of video-recordings does not provide more reliable analyses of speech disfluencies, and speech samples' characteristics and bilingual profile rather than SLPs' characteristics seem to influence the judgments. Qualitative analyses indicate that, in the current study, misidentification may be driven by neglecting or misinterpreting physical concomitants. In general, the findings indicate that identifying and analyzing speech fluency behaviors remain a challenging perceptual task, which underlies the need for developing consistent methods for training students and clinicians in identifying stuttering, especially in a bilingual context.