Paired-comparison hearing aid preferences: evaluation of an unforced-choice paradigm.

Fifteen normal-hearing listeners compared nine frequency-response slopes in a round-robin paired-comparison tournament as they listened to passages of connected discourse against a competing noise background at a +3-dB signal-to-noise ratio. All participants listened in an AB paradigm, in which they were forced to choose which of two signals (A or B) produced better speech intelligibility, and an ABN paradigm, in which they were allowed a third choice of No Preference (N). For both paradigms, listeners generally preferred frequency shaping that either cut low frequencies or boosted high frequencies, and intra- and intersession reliability was high overall. Although listeners' most-preferred responses tended to converge on the same frequency-response slope in the two paradigms, the reliability of these responses was significantly higher for ABN than for AB. The use of the ABN paired-comparison paradigm deserves further study, therefore, as a method to fit multimemory hearing aids.