Irrelevant speech effect in open plan offices: A laboratory study

Abstract: It seems now accepted that speech noise in open plan offices is the main source of discomfort for employees. This work follows a series of studies conducted at INRS France and INSA Lyon based on Hongisto's theoretical model (2005) linking the Decrease in Performance (DP) and the Speech Transmission Index (STI). This model predicts that for STI values between 0.7 and 1, which means a speech signal close to 100% of intelligibility, the DP remains constant at about 7%. The experiment that we carried out aimed to gather more information about the relation between DP and STI, varying the STI value up to 0.9. Fifty-five subjects between 25-59 years old participated in the experiment. First, some psychological parameters were observed in order to better characterize the inter-subjects variability. Then, subjects performed a Working-Memory (WM) task in silence and in four different sound conditions (STI from 0.25 to 0.9). This task was customized by an initial measure of mnemonic span so that two different cognitive loads (low/high) were equally defined for each subject around their span value. Subjects also subjectively evaluated their mental load and discomfort at the end of each WM task, for each noise condition. Results show a significant effect of the STI on the DP, the mental load and the discomfort. Furthermore, a significant correlation was found between the age of subjects and their performance during the WM task. This result was confirmed by a cluster analysis that enabled us to separate the subjects on two different groups, one group of younger and more efficient subjects and one group of older and less efficient subjects. General results did not show any increase of DP for the highest STI values, so the "plateau" hypothesis of Hongisto's model cannot be rejected on the basis of this experiment.