Impact of digital labeling on outcome measures

Abstract

Objective: The purpose of this study was to determine whether the label attached to the hearing aid being presented would bias outcome measures towards newer technological designs. Design: Two groups of subjects participated in this investigation. The groups were matched for age, gender, previous hearing aid experience, degree and configuration of hearing loss. Group A wore each of two digital hearing aids for 1 mo; Group B wore the same digital hearing aid for 2 mo, but the subjects were given the impression they were changing hearing aids after 1 mo. In each group the subjects were told that one of the months they were wearing a "digital" hearing aid and one of the months they were wearing a "conventional" hearing aid. Outcome measures consisted of a number of behavioral speech perception tasks and self-report measures, each completed at the onset and after 1 mo use with the hearing aids. Results: Labeling effects were observed for many of the outcome measures. Using a mixed-model factorial analysis of variance to control for irrelevant variables and to explore interaction terms, prejudice (digital versus conventional labeling) was treated as a within-subject factor while the subject group (A or B) and clinician were treated as between-subject factors. Although only the APHAB RV and BN scales showed significant labeling effects on their own, the group of tests used in this study showed a significant labeling effect as a whole (p < 0.01). The total influence of labeling and related interaction terms indicated labeling-related effects accounted for 2 to 32% of the variance in individual outcome measures. Discussion: The results of this investigation indicate a need for double-blinding in hearing aid research aimed at assessing the effectiveness of newer technologies, as well as a need for clinicians to critically evaluate the research describing the potential advantages of certain circuit options.