start » program » presentations

Text comprehension during noise exposure: Effects on eye movements, galvanic skin responses and subjective performance

In everyday urban life, noise and music is constantly present all around us, and there are few places where reading can be done in complete silence apart from studying in your own home, if even there. It is therefore necessary to further investigate how sounds tax cognition and facilitate or impede reading. Studies in acoustics and environmental medicine have shown evidence of cognitive effects of noise exposure. Furthermore, a previous study within this project showed that reading comprehension was negatively affected by music that the participant stated they would never use as study music, but not by music they preferred, café noise and silence. Crucially, the gaze behavior apart from pupil size was unaffected. The present study aims to investigate how reading behavior is influenced by noise presented at different sound pressure levels. In order to examine this, we collected eye movement data and galvanic skin response (GSR) for participants presented with multitalker babble noise presented at different levels during a reading comprehension task. Eye movement data were recorded binocularly with a tower-mounted SMI Hi-Speed eye tracker at 500Hz, and GSR with a BIOPAC system at 500Hz. The participants were presented with a babble noise at 50dB SPL and 70dB SPL, as well as a silent control condition, while reading texts with corresponding reading comprehension questions. The research questions are: do noise presentation level influence gaze behavior and other physiological responses during a text reading task; do stress levels (GSR) change with sound pressure levels (SPL), and how do stress levels affect eye movements and text comprehension; and does different sound pressure levels affect physiological process measures and cognitive product measures differently?

The 70 dB SPL noise condition was subjectively reported having a more negative effect on task performance than the 50 dB SPL noise condition, which in turn was reported having a more negative effect than the silent condition. However, the results showed no negative effect, and fixation duration, saccade amplitude, regressions, and second pass reading were largely unaffected. However, there was a significant effect on pupil size and blink rate. Furthermore, the GSR data indicated increased stress during the 70 dB SPL noise condition compared to the other ones, and the data suggested an interaction between noise level and test duration.

Findings suggest that noise does not affect text comprehension or higher level gaze behavior such as fixation duration, saccade amplitude and regressions. Crucially, low level cognitive systems such as stress levels (GSR), arousal (pupil size) and blink frequency were all influenced by sound pressure level. Therefore, future studies will examine the effect of different background noises on eye movements and GSR on a lower level than reading comprehension. Reading makes it difficult to isolate exactly what measures that are affected by the sound environment, and almost no results on reading behavior were found in the two previous experiments. Furthermore, participants reported the noise as being disturbing, and stated that they felt it worsened their performance, something that the product measures disprove. However, the GSR data suggest an effect of test duration. Future research should thus examine if even longer test durations affect performance.