An investigation of measurement equivalence in hearing response scales: refinement of a questionnaire for use in hearing screening.

PURPOSE

Questionnaires used in hearing screening should be short and demonstrate measurement equivalence across groups defined by hearing impairment and hearing aid experience. The measurement equivalence of two scales addressing 'functionality' (experienced hearing ability) and 'social hearing' (social barriers due to hearing problems) was investigated.

METHOD

Measurement equivalence was assessed using the Differential Item Functioning and Differential Test Framework within Item Response Theory. Three comparisons were considered: (1) Persons with normal hearing, reference group versus hearing impaired persons either owning or not owning a hearing aid; (2) Hearing aid users versus non-hearing aid users; (3) Hearing aid users versus hearing impaired persons without a hearing aid. A protocol of differential item detection was applied consisting of ordinal regression and the Log-likelihood ratio test to flag suspect items, followed by applying the log-likelihood ratio test using anchor items (items not suspect of differential item functioning).

RESULTS

The 11 item 'functionality' scale was reduced to 9 while the 10 item 'social hearing' scale was reduced to 7 items.

CONCLUSIONS

Applying the differential item functioning framework resulted in shorter questionnaires displaying measurement equivalence relative to hearing impairment and hearing aid use without loss of reliability.