Manchaiah, V., Oosthuizen, I., & Swanepoel, D.W.
American Auditory Society (AAS), Scottsdale, Arizona (USA), March 2023.
Publication year: 2023

Objectives: Hearing aid outcomes such as benefit and satisfaction are generally measured using structured outcome measures. However, not all items of these measures are relevant to all users and some items relevant to some users may not be captured in standardized measures. For this reason, open-ended questions are complementary to elicit responses that are highly relevant to individual users. The current study examined the psychologically meaningful dimensions using automated Natural Language Processing (NLP) techniques and compared it to a standardized hearing aid outcome measure.

Design: The study used a prospective cross-sectional survey design. Study participants (n=626) completed an online survey with demographic (e.g., age, gender) and audiological (e.g., duration of hearing loss, self-reported hearing difficulty) questions, open-ended questions about meaningful life changes following hearing aid use, and the International Outcome Inventory for Hearing Aids (IOI-HA). The responses to open-ended questions were analysed using the automated NLP software, Linguistic Inquiry Word Count (LIWC), which identifies psychologically meaningful language dimensions from the open-text response. The language dimensions were correlated with demographic and audiological variables as well as with the IOI-HA responses.

Results: The most frequently occurring language dimensions in the open-ended texts about meaningful life changes due to hearing aid use cognitive processes, social processes, I-words, motives, auditory perceptual processes, and affect (e.g., positive tones). LIWC dimensions had significant correlations mostly with residual activity limitations (IOI-HA Q3), residual participation restriction (IOI-HA Q5), and the IOI-HA total score. Age was significantly associated with the use of I-words, social referents, and lifestyle factors. Self-reported hearing difficulty and hearing loss duration were significantly correlated with cognitive processes, affect, and lifestyle factors.

Conclusions: LIWC helps identify meaningful psychological dimensions that are relevant to hearing aid users. The correlation between the LIWC variables and the IOI-HA supports the idea that the use of open-ended questions may complement the information obtained via the standardized hearing aid outcome and help optimize the outcome measurement in the clinical context.