Measuring the ability of children to understand everyday speech using computer technology: a normative study

Abstract

The present study aimed to standardize The University of Queensland Understanding Everyday Speech Test (UQUEST), a computer-based auditory test for measuring the ability of children to understand everyday speech, for Australian school-aged children. Participants were 41 Grade 3, and 58 Grade 4 students (55 boys/44 girls), attending primary schools in the Brisbane and Sunshine Coast areas. Their mean age was 8.3 years (SD = 0.7). The speech material, consisting of 20 test passages, was delivered to a child via headphones at 65 dBSPL with the concurrent presentation of video animations. The child listened to these passages under four different listening conditions (quiet, signal-to-noise (SNR] = 10 dB, 5 dB and 0 dB). After tlze presentation of each passage, the child answered four content-related multiple-choice questions. The results revealed different mean speech scores for the 20 passages for each listening condition. In general, very high mean scores for the passages were obtained for the quiet and 10 dB SNR noise conditions, indicating a possible ceiling effect. However, for the harder listening conditions (SNR = 0 and 5 dB), the mean passage scores were comparatively lower (59–96%). The present study suggested the use of six passages, which are of comparatively equal difficulty, and free from the effects of gender and education level, across the two harder test conditions for future testing. Normative data and critical differences, based on the scores of these six passages, were established.