Beyond Consequences of Challenge: the Degree to which Spoken Word Recognition Changes in Response to Difficulty has Independent Relevance

Listeners must often recognize words in challenging listening conditions like background noise. The cognitive mechanisms of spoken word recognition offer a potential locus for adaptation to such contexts. Indeed, research has found that the millisecond-by-millisecond dynamics of word recognition in challenging conditions differ lawfully along a small number of dimensions. However, comparatively little is known about the factors that impact the nature of an individual listener’s response to challenge. Here we ask if this response is stable across time and challenge type, if it is constrained by the idiosyncrasies of how listeners recognize words in quiet, and if these patterns of processing predict performance in real-world outcome measures. We tested listeners in a Visual World Paradigm with two different types of challenging listening across two days. We then leveraged a principal component analysis to extract underlying dimensions of variance across these conditions. Using this approach, we found that the dynamics of word recognition in challenge are predicted by their dynamics in typical speech, that these dynamics are stable across days and different types of challenge, and that the magnitude with which listeners change recognition predicts their performance in tests of speech-in-noise comprehension.