Copresence Was Found to Be Related to Some Pupil Measures in Persons With Hearing Loss While They Performed a Speech-in-Noise Task

Incorporating Virtual Reality Agents During a Dichotic Speech Reception Task: Insights From the Heart

OBJECTIVES

Listening effort is moderated by not only task difficulty, but also success importance. In real communication scenarios, success importance varies based upon the social context. However, in the laboratory, it can be challenging to manipulate social context without compromising experimental control. Outside of hearing sciences, studies have applied virtual reality (VR) to incorporate social context in a controlled and repeatable manner. Several of these studies have demonstrated that social manipulations in VR can reliably elicit changes in cardiovascular measures. Here, we investigated the effect of adding VR agents to a speech reception task, while measuring from the cardiovascular system.

DESIGN

Twenty-eight, normally hearing participants undertook a dichotic speech reception task. Sentences in stationary noise were presented dichotically, that is, different sentences presented simultaneously to each ear. Participants were tasked to either repeat one of the sentences (single-sentence condition) or both of the sentences (dual-sentence condition). The task was conducted under two VR conditions: (1) in the presence of agents, who provided sporadic performance feedback and (2) in the presence of nonagent controls, without any feedback given. Alongside task performance, we quantified changes in cardiovascular measures, relative to pretask baselines: heart rate variability, pre-ejection period, heart rate, and blood pressure. After each condition, participants rated their subjective effort, difficulty, performance, and engagement.

RESULTS

Performance and the subjective perception of performance were lower, while subjective effort and difficulty were higher, in the dual-sentence condition, compared with the single-sentence condition. Heart rate was the only cardiovascular measure that was sensitive to the experimental manipulations. Contrary to our expectations, heart rate increased in the nonagent control conditions, compared with the agent conditions. An exploratory analysis revealed heart rate fluctuations within a trial: heart rate was higher during the first 6 sec of the trial (reflecting the presentence masking noise and the sentence presentation) in the dual-sentence condition, compared with the single-sentence condition.

CONCLUSIONS

This study was the first to incorporate VR agents who provided performance feedback during a dichotic speech reception task. Our results suggest that the VR agents did not increase success importance, which could be attributed to a lack of realism of the agents. We also demonstrated that the cardiovascular response to experimental manipulations may differ depending on the data window selected for analysis.

Combining Cardiovascular and Pupil Features Using k-Nearest Neighbor Classifiers to Assess Task Demand, Social Context, and Sentence Accuracy During Listening

In daily life, both acoustic factors and social context can affect listening effort investment. In laboratory settings, information about listening effort has been deduced from pupil and cardiovascular responses independently. The extent to which these measures can jointly predict listening-related factors is unknown. Here we combined pupil and cardiovascular features to predict acoustic and contextual aspects of speech perception. Data were collected from 29 adults (mean  =  64.6 years, SD  =  9.2) with hearing loss. Participants performed a speech perception task at two individualized signal-to-noise ratios (corresponding to 50% and 80% of sentences correct) and in two social contexts (the presence and absence of two observers). Seven features were extracted per trial: baseline pupil size, peak pupil dilation, mean pupil dilation, interbeat interval, blood volume pulse amplitude, pre-ejection period and pulse arrival time. These features were used to train k-nearest neighbor classifiers to predict task demand, social context and sentence accuracy. The k-fold cross validation on the group-level data revealed above-chance classification accuracies: task demand, 64.4%; social context, 78.3%; and sentence accuracy, 55.1%. However, classification accuracies diminished when the classifiers were trained and tested on data from different participants. Individually trained classifiers (one per participant) performed better than group-level classifiers: 71.7% (SD  =  10.2) for task demand, 88.0% (SD  =  7.5) for social context, and 60.0% (SD  =  13.1) for sentence accuracy. We demonstrated that classifiers trained on group-level physiological data to predict aspects of speech perception generalized poorly to novel participants. Individually calibrated classifiers hold more promise for future applications.

The Influence of Hearing Loss on the Pupil Response to Degraded Speech

PURPOSE

Current evidence regarding the influence of hearing loss on the pupil response elicited by speech perception is inconsistent. This might be partially due to confounding effects of age. This study aimed to compare pupil responses in age-matched groups of normal-hearing (NH) and hard of hearing (HH) listeners during listening to speech.

METHOD

We tested the baseline pupil size and mean and peak pupil dilation response of 17 NH participants (Mage = 46 years; age range: 20-62 years) and 17 HH participants (Mage = 45 years; age range: 20-63 years) who were pairwise matched on age and educational level. Participants performed three speech perception tasks at a 50% intelligibility level: noise-vocoded speech and speech masked with either stationary noise or interfering speech. They also listened to speech presented in quiet.

RESULTS

Hearing loss was associated with poorer speech perception, except for noise-vocoded speech. In contrast to NH participants, performance of HH participants did not improve across trials for the interfering speech condition, and it decreased for speech in stationary noise. HH participants had a smaller mean pupil dilation in degraded speech conditions compared to NH participants, but not for speech in quiet. They also had a steeper decline in the baseline pupil size across trials. The baseline pupil size was smaller for noise-vocoded speech as compared to the other conditions. The normalized data showed an additional group effect on the baseline pupil response.

CONCLUSIONS

Hearing loss is associated with a smaller pupil response and steeper decline in baseline pupil size during the perception of degraded speech. This suggests difficulties of the HH participants to sustain their effort investment and performance across the test session.