Abstract |
Purpose
The purpose of this study was to investigate the effect of contrast, print size, and luminance as well as the learning effect, on eye movement behavior when reading. The impact on the distribution of fixation duration was investigated using eye-movement recording and ex-Gaussian analysis.
Methods
A group of 23 persons (age; 39 -45 years) participated in the study. Two experiments were conducted. The first experiment investigated the effect of contrast and print size on reading as well as the learning effect. The second experiment evaluated the effects of luminance on reading. Two types of reading cards were used in the study; the Greek versions of the Colenbrander Reading Cards (both normal and mixed contrast) and a paragraph of fixed print size, the “long text”. The eye movement recording was performed at a distance of 40 cm using the eye tracker EyeLink II by SR Research Ltd.
Results
No statistical significant difference was found in median fixation duration between the first and the second time of reading the text at all print sizes. Mean number of fixations per sentence was found to decrease by about one fixation in the second time of reading in all print sizes. Median fixation duration was significantly affected by print size when letters were smaller than 0.4 logMAR and the effect was more pronounced as letters reduced further in size. Contrast influenced fixation duration in a statistically significant matter when letters were smaller than 0.4 logMAR (p=0.023). In contrary, luminance had a significant effect on median fixation duration at all letter sizes (at 0.9 logMAR, p=0.014). The change on fixation duration in low luminance is getting more evident in 0.5 logMAR print size and for smaller letters. Using the ex-Gaussian analysis and a print size of 0.4 logMAR we found that the
fixation duration distribution was identical between the first and the second time of reading the same text (p=0.181; Kolmogorov-Smirnov test). On the contrary, the contrast had a significant effect on the distribution (p=0.009; Kolmogorov-Smirnov test), which means that low contrast distribution is different and with smaller values that the high contrast distribution (p=0.006; Kolmogorov-Smirnov test).
Conclusions
The ex-Gaussian analysis was proved to be a very precious tool for understanding eye movement behavior when reading. We found that optical factors, like luminance and contrast affect in the fixation duration distribution. Ex-Gaussian parameters give a quantitative view of great importance.
To our opinion, the combination of standardized long texts with a proper analyzing method, like the one that we followed, will lead to a much better understanding on the effect of several factors on reading. Moreover, further studies can work on distinguishing the optical effects (contrast, luminance, print size, aberrations etc.) from cognitive (dyslexia, understanding problems etc.), psychological and neurological factors (anxiety, autism, etc.)
|