Test-Retest Reliability of Ecological Momentary Assessment in Audiology Research
Yu-Hsiang Wu1, Elizabeth Stangl1, Octav Chipara2
1Department of Communication Sciences and Disorders, The University of Iowa, Iowa City, Iowa.
Background:
Ecological momentary assessment (EMA) is a methodology involving repeated surveys to collect in situ data that describe respondents' current or recent experiences and related contexts in their natural environments. Audiology literature investigating the test-retest reliability of EMA is scarce.
Purpose:
This article examines the test-retest reliability of EMA in measuring the characteristics of listening contexts and listening experiences.
Research Design:
An observational study.
Study Sample:
Fifty-one older adults with hearing loss.
Data Collection And Analysis:
The study was part of a larger study that examined the effect of hearing aid technologies. The larger study had four trial conditions and outcome was measured using a smartphone-based EMA system. After completing the four trial conditions, participants repeated one of the conditions to examine the EMA test-retest reliability. The EMA surveys contained questions that assessed listening context characteristics including talker familiarity, talker location, and noise location, as well as listening experiences including speech understanding, listening effort, loudness satisfaction, and hearing aid satisfaction. The data from multiple EMA surveys collected by each participant were aggregated in each of the test and retest conditions. Test-retest correlation on the aggregated data was then calculated for each EMA survey question to determine the reliability of EMA.
Results:
At the group level, listening context characteristics and listening experience did not change between the test and retest conditions. The test-retest correlation varied across the EMA questions, with the highest being the questions that assessed talker location (median r = 1.0), reverberation (r = 0.89), and speech understanding (r = 0.85), and the lowest being the items that quantified noise location (median r = 0.63), talker familiarity (r = 0.46), listening effort (r = 0.61), loudness satisfaction (r = 0.60), and hearing aid satisfaction (r = 0.61).
Conclusion:
Several EMA questions yielded appropriate test-retest reliability results. The lower test-retest correlations for some EMA survey questions were likely due to fewer surveys completed by participants and poorly designed questions. Therefore, the present study stresses the importance of using validated questions in EMA. With sufficient numbers of surveys completed by respondents and with appropriately designed survey questions, EMA could have reasonable test-retest reliability in audiology research.
