Development of a subjective scale for sound quality assessments in building acoustics

Abstract

In building acoustics, it is necessary to conduct both objective and subjective evaluations in a holistic sound quality assessment. However, there is a huge variety in a selection of psychological tools. This arouses the concern about reliability and validity of the tools. Although various perceptions can be affected by environmental sounds, the underlying structure of acoustic perceptual influences was recently investigated by the researchers in their systematic review study. This is a first study aiming at the development of a valid and reliable subjective scale to quantitatively assess the three fundamental perceptual dimensions of sound via the investigation of its psychometric properties. Nine semantic differential questions in the psychoacoustics perception scale (PPS) covered the assessments about the subjects' responses to the general judgement (Evaluation (E)), energy content (Potency (P)), and temporal and spectral content (Activity (A)) of sound. The reliability test results (Cronbach's αs > 0.80) indicated the acceptable internal consistencies of the items in the E, P, and A factors. The construct validity in characterizing the structure of perceptual influences was confirmed by the goodness-of-model-fit indexes in the confirmatory factor analysis (CFA, N = 128) for the factorial structure of the hypothetical EPA model behind the PPS. The further invariant tests verified the invariances of the model across gender except the error variance of the E factor. A total EPA score, representing the joint attribution of the factors, was a significant predictor of the other perceptions. The concurrent validity of the PPS to the modified dental anxiety scale (MDAS), a well-developed psychometric tool, was also supported by the result that the E factor was a significant predictor of the MDAS score.