Application of Paired-Comparison Methods to Hearing Aids

Abstract

The method of paired comparisons was introduced into the hearing aid literature nearly 50 years ago. Over time, studies have found paired comparisons to be sensitive, valid, and reliable in determining either the perceptual difference or relative ranking among hearing aids and electroacoustic characteristics. With the increasing number of adjustable electroacoustic parameters in today's digital hearing aids-and the lack of procedural guidelines necessary to fit many of them-the method of paired comparisons provides the clinician with the ability to compare different devices, electroacoustic characteristics, memory settings, or combinations of these variables under the listener's everyday listening conditions. Furthermore, this procedure provides the clinician with the ability to individualize the prescriptive approach-which is predicated mainly on hearing threshold data and listening in quiet-so that a combination of parameters can be set to optimize the user's listening needs in a given environment. In this article, the authors present an overview of the theoretical principle supporting this procedure, the various paired-comparison strategies and associated approaches, the advantages of this method, and recommended procedures for implementing the method of paired comparisons in the fitting of today's sophisticated hearing aids.

Common Sound Scenarios: A Context-Driven Categorization of Everyday Sound Environments for Application in Hearing-Device Research

“…Another dimension commonly evaluated in the laboratory is sound quality, which can be measured using a large number of attributes, such as preference, speech clarity, comfort, and naturalness. Sound quality is often measured using scaling or paired comparisons (Gabrielsson and Sjögren, 1979;Eisenberg et al, 1997;Amlani and Schafer, 2009). The measurements are sensitive to small differences in signal processing.…”

Self-Adjusted Amplification Parameters Produce Large Between-Subject Variability and Preserve Speech Intelligibility

“…Preferred gain-frequency responses for hearing aids have previously been investigated using a variety of paradigms. Comparison or rating methods require listeners to make judgments about their preference for or perception of sounds after amplification with different gain-frequency responses, either as paired comparisons ( Amlani & Schafer, 2009 ; Byrne, 1986 ; Keidser, et al., 2005 ; Keidser, Dillon, & Byrne, 1995 , Kuk, Harper, & Doubek, 1994 ; Kuk & Lau, 1995a , 1995b ; Kuk & Lau, 1996b ; Kuk & Pape, 1992 , 1993; Moore, Füllgrabe, & Stone, 2011 ; Neuman, Levitt, Mills, & Schwander, 1987 ; Preminger, Neuman, Bakke, Walters, & Levitt, 2000 ; Punch & Howard, 1978 ; Punch & Parker, 1981 ; Punch, Rakerd, & Amlani, 2001 ; Smeds, 2004 ; Stelmachowicz, Lewis, & Carney, 1994 ) or individual, unpaired ratings ( Kuk & Lau, 1996a ; van Buuren, Festen, & Plomp, 1995 ). Another approach is to use adjustment methods which entail assessing the output of a hearing aid after it has been adjusted—often using the volume control—to better match the preferred listening level ( Boothroyd & Mackersie, 2017 ; Boymans & Dreschler, 2012 ; Cox & Alexander, 1991 , 1992; Dreschler, Keidser, Convery, & Dillon, 2008 ; Hornsby & Mueller, 2008 ; Horwitz & Turner, 1997; Humes,Wilson, Barlow, & Garner, 2002 ; Keidser, Dillon, & Convery, 2008 ; Marriage, Moore, & Alcántara, 2004; Polonenko et al., 2010 ; Smeds et al., 2006 ; Souza & Kitch, 2001) or by analyzing the output of trainable hearing aids after completion of a training regime ( Keidser & Alamudi, 2013 ; Mueller, Hornsby, & Weber, 2008 ; Zakis, Dillon, & McDermott, 2007 ).…”

Feasibility of Real-Time Selection of Frequency Tables in an Acoustic Simulation of a Cochlear Implant

“…To date there have been very few efforts directly addressing this issue, although there are two procedures that have been previously investigated which could theoretically be used to do so. These procedures are the simplex procedure (e.g., Neuman et al, 1987; Preminger et al, 2000; Franck et al, 2004; Amlani and Shafer, 2009) and the genetic algorithm (e.g., Holland, 1975; Wakefield et al, 2005; Baskent and Edwards, 2007; Baskent et al, 2007 ).…”