Listening in complex acoustic environments with degraded speech signals poses
considerable challenges, particularly for older adults with hearing loss. While these difficulties are
primarily attributed to elevated hearing thresholds, the involvement of cognitive processes,
especially working memory, in comprehending speech under suboptimal listening conditions is
well established. However, research on the specific contributions of various working memory span
tasks to the ability of older adults with hearing loss to recognize different types of degraded speech
has been inconsistent.
This dissertation examines the relationship between various working memory span tasks
(including the reading span test, list sorting task, and word/nonword repetition span) and the ability
of 28 older listeners (aged 61–87) with hearing loss to recognize three types of degraded speech:
reverberant speech, speech embedded in multi-talker babble noise, and time-compressed speech.
To standardize recognition task scores, we developed an adaptive procedure to establish the speech
reception threshold for each degraded condition.
Analysis revealed notable differences in group and individual performance across the
speech recognition tasks, necessitating careful determination of the SRT for each. Additionally, in
all tasks, participants showed significant improvement throughout, suggesting procedural learning
or rapid perceptual adaptation that may influence adaptive SRT measurement. Among working
memory tasks, only the reading span test significantly correlated with performance in both speechin-noise and time-compressed speech tasks and was a significant predictor of variance in the
speech-in-noise task. This implies that selecting specific working memory tasks, such as the
xv
reading span test, is essential when examining cognitive contributions to degraded speech
recognition.
These findings highlight the pivotal role of working memory in supporting speech
recognition under adverse conditions for older adults with hearing loss. The study underscores the
importance of incorporating cognitive factors, such as working memory capacity, into auditory
rehabilitation strategies, providing valuable insights into the interaction between cognitive and
auditory processes in this population. Furthermore, this work lays a foundation for advancing
adaptive procedures to assess speech reception thresholds across other types of degraded speech
beyond the established procedures for assessing the signal-to-noise ratio for 50% recognition
(SNR50) of speech in background noise.
Keywords: working memory capacity, reverberant speech, speech-in-noise, time-compressed
speech, older adults, hearing loss |