Methodology for the subjective evaluation of airborne sound insulation through 2-AC and Thurstonian models

Abstract

Single-Number Quantities (SNQs) are used to evaluate the performance of sound insulation elements through a global and quantitative value that takes into account the contribution of all frequency bands. In recent years, several studies have been carried out to investigate how representative these magnitudes are of the subjective perception of sound protection perceived by human beings. These studies have mainly been carried out through listening tests, in which a series of participants had to judge how they perceived the effect that certain sound insulation elements had on different sound samples, regarding several terms such as annoyance or loudness. Different methodologies have been used, in which possible sources of bias that may affect the results have been considered in varying degrees of detail, depending on the study. In addition, most of these studies were only able to evaluate the correlation between the usual SNQs and the subjective perception of sound insulation, but they were not able to determine the causes that have generated these correlations. This limitation has been mainly caused by the methods used for the design and analysis. The purpose of this article is to propose a detailed methodology, based on difference testing methods and Thurstonian models. This methodology focuses not only on evaluating the existing relationships between the different SNQs and the subjective perception of sound insulation but also on determining the causes that generate those relationships. The fact that this methodology focuses on determining the causes can help to identify the shortcomings of the current SNQs and propose new SNQs, which are more representative of the subjective perception of sound insulation. In order to address this proposal, all stages of the design, performance and analysis of the listening test are first described in detail, emphasizing the factors that need to be addressed with utmost care, as they may have a representative influence on the quality and relevance of the results. In addition, the advantages of carrying out an experimental design based on difference testing methods and Thurstonian models are presented. This methodology is then explained in detail by means of a case study, in which the subjective perception of different façade elements is evaluated. Through this case study, the benefits of an experimental design based on a 2-AC (2 alternative choice) difference testing method and a Thurstonian modeling approach for the analysis of the results are highlighted as an alternative for current methods of design and analysis.