Assessing the effect of pupillometric fatigue on task-evoked responses in a listening effort paradigm

Measurement of task-evoked pupillary responses (TEPR) is widely used for quantifying the cognitive status of participants (e.g., listening effort; LE) in hearing sciences. A key methodological challenge of pupillometry is that TEPR weakens over the course of the experimental session, raising questions of whether such fatigue influences participant response to experimental manipulations in later trials. To address this issue, we measured TEPR in a sentence recognition task where participants listened to, and repeated back, sentences in the presence of a distracting same-sex talker. To manipulate LE, each of the 200 trials was presented at one of 4 signal-to-noise ratios (SNR; −10 to +5 dB), varying pseudo-randomly on a trial-by-trial basis. Linear mixed-effects modeling of TEPR was used to assess whether chronological points in the experiment interacted with the effect of SNR. Preliminary analyses of peak TEPR suggest that despite substantial accrual of pupillometric fatigue over the experimental session, fatigue effects did not significantly interact with SNR, indicating that effects of LE manipulations on TEPR remain comparable across different levels of pupillometric fatigue. These results indicate TEPR may be a suitable measure of LE even in relatively long-lasting (e.g., ∼2 h) paradigms. [Research supported by NIH grant R21DC020788 and UMN’s UROP program.]