Purpose This study assessed state anxiety as a function of speech recognition testing using three clinical measures of speech in noise and one clinical measure of dichotic speech recognition. Method Thirty young adults, 30 middle-age adults, and 25 older adults participated. State anxiety was measured pre– and post–speech recognition testing using the State–Trait Anxiety Inventory. Speech recognition was measured with the Revised Speech Perception in Noise Test, the Quick Speech-in-Noise Test, the Words-in-Noise Test, and the Dichotic Digits Test (DDT). Results Speech recognition performance was as expected: Older adults performed significantly poorer on all measures as compared to the young adults and significantly poorer on the Revised Speech Perception in Noise Test, the Quick Speech-in-Noise Test, and the Words-in-Noise Test as compared to the middle-age adults. On average, State–Trait Anxiety Inventory scores increased posttesting, with the middle-age adults exhibiting significantly greater increases in state anxiety as compared to the young and older adults. Increases in state anxiety were significantly greater for the DDT relative to the speech-in-noise tests for the middle-age adults only. Poorer DDT recognition performance was associated with higher levels of state anxiety. Conclusions Increases in state anxiety were observed after speech-in-noise and dichotic listening testing for all groups, with significant increases seen for the young and middle-age adults. Although the exact mechanisms could not be determined, multiple factors likely influenced the observed increases in state anxiety, including task difficulty, individual proficiency, and age.

Few studies have directly compared adults' or children's perception of nonnative accents and unfamiliar regional dialects. However, some evidence suggests that nonnative varieties cause greater decrements in intelligibility and processing than unfamiliar native dialects, while metalinguistic awareness for nonnative varieties develops earlier than awareness for regional variants. To directly examine regional and nonnative accent perception, we tested sentence recognition in American-English monolingual 5- to 7-year-old children and adults for three accents: Central Midland (familiar native), Scottish (unfamiliar native), and German (detectable but mild nonnative accent) in quiet and multitalker babble. In quiet, both children and adults showed highly accurate word recognition for all accents. Although children's performance was lower than adults’ in noise, overall word recognition accuracy patterns across accents were similar: accuracy was highest for the Midland talker, followed by the German-accented talker, and poorest for the Scottish talker. These data suggest that the greater decrements for nonnative accents compared to unfamiliar regional dialects previously reported may have arisen from the specific varieties or talkers selected. Although both types of unfamiliar speech can cause listening difficulty in noisy environments, the acoustic-phonetic distance from the home dialect may predict both adults' and children's performance better than native vs. nonnative status.