Developmental effects in children’s ability to benefit from F₀ differences between target and masker speech

Objectives: The objectives of this study were to (1) evaluate the extent to which school-age children benefit from fundamental frequency (F₀) differences between target words and competing two-talker speech, and (2) assess whether this benefit changes with age. It was predicted that while children would be more susceptible to speech-in-speech masking compared to adults, they would benefit from differences in F₀ between target and masker speech. A second experiment was conducted to evaluate the relationship between frequency discrimination thresholds and the ability to benefit from target/masker differences in F₀. Design: Listeners were children (5 to 15 years) and adults (20 to 36 years) with normal hearing. In the first experiment, speech reception thresholds (SRTs) for disyllabic words were measured in a continuous, 60-dB SPL two-talker speech masker. The same male talker produced both the target and masker speech (average F₀ = 120 Hz). The level of the target words was adaptively varied to estimate the level associated with 71% correct identification. The procedure was a four-alternative forced-choice with a picture-pointing response. Target words either had the same mean F₀ as the masker or it was shifted up by 3, 6, or 9 semitones. To determine the benefit of target/masker F₀ separation on word recognition, masking release was computed by subtracting thresholds in each shifted-F₀ condition from the threshold in the unshifted-F₀ condition. In the second experiment, frequency discrimination thresholds were collected for a subset of listeners to determine whether sensitivity to F₀ differences would be predictive of SRTs. The standard was the syllable /ba/ with an F₀ of 250 Hz; the target stimuli had a higher F₀. Discrimination thresholds were measured using a three-alternative, three-interval forced choice procedure. Results: Younger children (5 to 12 years) had significantly poorer SRTs than older children (13 to 15 years) and adults in the unshifted-F₀ condition. The benefit of F₀ separations generally increased with increasing child age and magnitude of target/masker F₀ separation. For 5- to 7-year-olds, there was a small benefit of F₀ separation in the 9-semitone condition only. For 8- to 12-year-olds, there was a benefit from both 6- and 9-semitone separations, but to a lesser degree than what was observed for older children (13 to 15 years) and adults, who showed a substantial benefit in the 6- and 9-semitone conditions. Examination of individual data found that children younger than 7 years of age did not benefit from any of the F₀ separations tested. Results for the frequency discrimination task indicated that, while there was a trend for improved thresholds with increasing age, these thresholds were not predictive of the ability to use F₀ differences in the speech-in-speech recognition task after controlling for age. Conclusions: The overall pattern of results suggests that children’s ability to benefit from F₀ differences in speech-in-speech recognition follows a prolonged developmental trajectory. Younger children are less able to capitalize on differences in F₀ between target and masker speech. The extent to which individual children benefitted from target/masker F₀ differences was not associated with their frequency discrimination thresholds.