PP1332 - The Effect of Beat Sensitivity Determined by a Covert Pulse Continuation Task as It Relates to Working Memory and Speech Recognition in Noise Ability

Objective:
The purpose of this study was to determine the relationship between beat sensitivity vs. working memory and speech-in-noise ability.  

Rationale:  

Cinelyte et al., (2022) developed a measure of internal-timekeeping mechanisms, an aspect of beat perception which is defined as the covert continuation of an internal pulse. The Beat-Drop-Alignment Test (BDAT) assesses a participant's ability to identify the beat in a musical excerpt and judge whether a single probe falls on or off the beat. This is a measure of beat sensitivity. Cinelyte reported significantly better BDAT performance for musicians than non-musicians. Parbery-Clark et al., (2009) demonstrated that musicians had better working memory and speech in noise ability than non-musicians (p = 0.041). They also reported a statistically significant correlation between working memory vs. speech-in-noise ability. Three hypotheses were made: 1) There will be a statistically significant relationship between beat sensitivity vs. speech-in-noise ability. 2) There will be a statistically significant relationship between working memory vs. speech-in noise ability. 3) There will be a statistically significant relationship between working memory vs. beat sensitivity. 

Design:  

Thirty-one native English speakers participated in the study (mean age = 21.09, SD = 2.7). Pure-tone thresholds were within normal limits (≤ 25 dB HL, 0.25 – 6.0 kHz) for all participants. The standard adaptive HINT protocol was administered for the Noise Front condition under headphones. The Noise Front threshold is the signal-to-noise ratio (SNR) where the participant recognizes 50% of the sentences. The HINT sentences were presented in steady-state-speech-shaped noise presented at 65 dBA. The Az Bio sentences, and a ten-talker babble were presented at 65 dBA (0 dB SNR). Ten out of the original fifteen Az Bio sentence lists were used in this study. Beat sensitivity was determined using an adaptive item response model from -4 (best score) to +4 (worst score). The Beat Drop stimuli (music and probe) was presented at a 55 dB dial setting. This level was found to be a comfortable listening level during a pilot study. Working memory was determined using the Digit Span task, the stimuli was presented at 65 dBA.  

Results:  

No statistically significant relationship was found between beat sensitivity vs. HINT thresholds (r = -0.0012, p = 0.9948). No statistically significant relationship was found between working memory vs. HINT thresholds (r = -0.1844, p = 0.3206). A statistically significant relationship was found between working memory vs. beat sensitivity (r = 0.3714, p = 0.0397), indicating a strong and weak convergent task validity. Mean score for HINT Steady State Noise Front Threshold, Az Bio-10-talker score, Digit Span score, and BDT score, were 0.17%, 53.81%, 5.85%, and -0.89%, respectively. The range for the HINT Threshold, Az Bio-10-talker score, Digit Span score, and BDT score, were 8.9, 88.73, 9, and 5.07, respectively. The maximum score for HINT Steady State Noise Front Threshold, Az Bio-10-talker score, Digit Span score, and BDT score, were 6.7%, 93.08%, 12%, and 2.11% for all subjects.  

Conclusions:  

A statistically significant relationship was found between beat sensitivity vs. working memory. This indicates that working memory capacity and beat sensitivity predictability provide evidence of convergent validity.