The intelligibility and the temporal characteristics of background sounds: Which impact do they have on cognitive performance?

Working in open-plan offices implies performing mental task during background speech. Many studies have investigated the negative effects of background speech on verbal short-term memory, well known as the Irrelevant Sound Effect (ISE). Irrelevant background sounds with distinct temporal-spectral variations induce this empirically robust phenomenon. Although non-speech sounds with these characteristics elicit the ISE, the detrimental effect of background speech is unsurpassed. Its disturbance effect has been shown to vary as a function of the background speech´s intelligibility: reduced intelligibility results in less detrimental effects. The present experiment explores the question whether the intelligibility of background speech or its temporal-spectral variations accounts for detrimental effects on cognitive performance. Overall, 24 subjects were tested in eight sound conditions in a serial recall task, which is the standard task for testing short-term memory. Sound conditions differed with respect to instrumentally measured speech intelligibility (Speech Transmission Index, STI) and temporal-spectral variations (hearing sensation fluctuation strength). Perfectly intelligible speech reduced short-term memory significantly more compared to a non-speech sound of equal fluctuation strength. Step-wise reduction of speech intelligibility, accomplished by masking the speech signal with continuous noise of different signal-to-noise ratio (SNR = 0, 3 or 6 dB(A)), significantly reduced the error rates. The experiment indicates that the intelligibility of speech rather than its temporalspectral characteristics accounts for a greater portion of the disturbance effect. Thus, this study underlines once more the necessity of reducing background speech´s intelligibility in open-plan offices.