Matching laboratory and real-world listening environments with ecological momentary assessment improves prediction of speech perception.

Abstract

Laboratory speech-in-noise measures fail to predict real-world listening performance. This study tested whether matching laboratory to real-world environments using self-reported noise sources from ecological momentary assessment (EMA) improves predictions. Speech intelligibility and listening effort were measured in virtual acoustic environments in 51 normal-hearing adults, who then self-rated both via EMA for 1 week. Laboratory intelligibility predicted real-world intelligibility only when laboratory and real-world environments were matched. Effort results were not significant. Findings suggest that environment-specific laboratory testing improves ecological validity of intelligibility measures, though environment matching explained only 1.1% of variance in real-world ratings, with individual differences accounting for substantially more.