Monitoring reading behaviour: examining eye metrics during processing of information with different levels of relevance

• ¹ Netherlands Organisation for Applied Scientific Research (TNO), Netherlands

The technological improvements in eye tracking allow addressing novel and innovative questions in eye movement behaviour and its relation to cognition in real-life conditions. Multiple studies reveal for instance, how pupil size reflects mental effort, how blink rate is involved in learning and goal-oriented behaviour, and how gaze patterns are associated with current attentional focus. Mental state monitoring through eye tracking can be useful for a wide area of applications. In this project, an experiment was performed in order to examine several eye metrics while reading two information boxes, each on one of two topics, with different levels of relevance (figure 1). Questions on the read information were asked afterwards. In order to create a difference in relevance, 80 percent of the questions were solely about one of the two topics, indicated by a small green outline. Before starting the task, participants were informed which topic this would be. Visit duration, fixation duration, pupil size and blink rate were compared between reading the relevant and less relevant information box. It was found that visit duration and average fixation duration were increased when reading text with higher relevance (figure 2), while blink rate, blink duration and average pupil size did not differ. These results implicate that even though blink rate and pupil size have been found to reflect attention and mental effort before, visit duration and fixation duration are relatively more useful markers of relevance when reading information in a life-like setting. Figure 1. Trial design of the Information Processing task. A stimulus slide contains a ‘relevant information box’ and a ‘irrelevant information box’. Stimulus slides are separated by a screen with a fixation cross. After two stimulus slides five multiple choice questions are asked, four focusing on the information in the relevant boxes, one on the information in the irrelevant boxes. This procedure was repeated for 20 trials. Figure 2. Visit duration and fixation duration for the relevant and less relevant information boxes. Error bars represent the standard error of the mean (SEM).