Serial Recall Predicts Vocoded Sentence Recognition Across Spectral Resolutions.

Only 32 of 38 recruited participants were analysed for the main comparison, and the sample was young adults with typical hearing and normal or corrected-to-normal vision. The study did not include people with hearing loss, older adults, or clinical populations, so the relationships may not apply to patients. The speech task used a single vocoded stream without competing noise, which likely reduced attention demands. The study was not preregistered, and some analyses, including the composite score and reaction-time distribution work, were post hoc. The fluid intelligence scores were positively skewed relative to norms, which can make relationships harder to detect. The vocoder-sensitivity component was small, so individual estimates of it were imprecise. Serial recall is not a pure measure of one memory ability; it reflects a mixture of processes.

This work was supported by an NIH COBRE grant awarded to Adam Bosen. Human Subjects Core staff recruited participants and administered hearing, vision, and WASI testing. The Test of Variables of Attention Company provided discounted test credits. David Pisoni provided PRESTO recordings, Elizabeth Schneider spoke the stimuli, and Mary Luckasen helped design the word list.

Do not conclude that serial recall is a validated clinical predictor for people with hearing loss. The participants were young adults with typical hearing, and the study reports correlations in vocoded sentence recognition, not evidence that the relationships apply to people with hearing loss.