Abstract:
The increased popularity of personal listening devices has given rise to concerns relating to the potential for hearing damage caused by their use. It is well established that increased background noise causes individuals to increase the output of their personal listening devices; however it remains unclear whether some individuals are more affected by background noise than others. Anecdotal evidence suggests some variability in preferred listening levels in noise may be related to the ability to recognise speech in noise. In the present study the preferred listening levels of 33 participants were recorded using a real ear measurement setup while they listened to a pre-selected podcast in quiet and in four different intensities of babble noise and speechweighted noise. Preferred listening levels were compared to several primary predictor measures including QuickSIN score, acceptable noise level, OAE suppression, response to questions relating to hearing in noise, standard- and high- frequency audiometric average, and preferred listening level in quiet. The results revealed a number of measures, including QuickSIN score, predicted preferred listening levels in background noise. Linear regression was used to model the effect of different primary predictors on signal-to-noise ratios in background noise and a number of significant models emerged. Preferred listening level in quiet, QuickSIN score and audiometric average appeared in the most models predicting signal-to-noise ratio in both babble noise and speech weighted noise. The models imply multiple factors, including the ability to recognise speech in noise, influence listening behaviours in noise. Furthermore, the influence of different factors may change as a function of background noise intensity.
