# Comparison of Test–Retest Reliability of Sound Field Audiometry Between a Newly Designed System for Small Audiometric Booths and a Conventional Sound Field System

## TL;DR

This study compares a new sound field audiometry system for small booths with a conventional system, finding similar reliability and better word recognition scores in the new system.

## Contribution

A newly designed sound field system for small booths is shown to be equivalent to conventional systems in reliability and performance.

## Key findings

- Aided warble-tone thresholds and speech reception thresholds showed no significant test–retest differences in both systems.
- The new system produced higher aided word recognition scores at 65 dB HL compared to the conventional system.
- Warble-tone thresholds between the two systems were equivalent within a ±10 dB margin.

## Abstract

Objective: Sound field (SF) audiometry is widely used to evaluate aided hearing performance. This study compared the test–retest reliability and between-system equivalence of SF audiometry between a newly designed SF system for small audiometric booths and a conventional SF system. Methods: Thirty-nine adults using hearing aids (56 tested ears; 19 females [26 ears] and 20 males [30 ears]) underwent aided warble-tone audiometry (0.25, 0.5, 1, 2, and 4 kHz) and aided speech audiometry using a conventional SF system (two loudspeakers at 45° azimuth and 1 m distance) and a newly designed small-booth SF system (two height-adjustable loudspeakers at 45° azimuth and at 30 cm distance). The same sequence was repeated to assess test–retest performance. The test–retest differences were evaluated using paired t-tests, and the variability was summarized using the coefficient of variation (CV). Equivalence of aided warble-tone thresholds between systems was evaluated using two one-sided tests (TOST) with a ±10 dB margin. Results: Within each system, aided warble-tone thresholds and aided speech reception thresholds (SRTs) did not differ significantly between the test and the retest (all p > 0.05). The warble-tone thresholds were equivalent between systems within the predefined ±10 dB margin by TOST. Aided word recognition scores (WRSs) at 65 dB HL were higher in the newly designed system than in the conventional system (p < 0.0001). The CVs were low in both systems and were slightly lower in the newly designed system. Conclusions: When the listener-to-loudspeaker position is strictly controlled, SF audiometry provides stable test–retest results in hearing-aid users. The newly designed SF system for small audiometric booths produced aided thresholds equivalent to those of the conventional SF system and yielded higher WRSs at 65 dB HL under the tested conditions.