The Effects of Typographic Features on the Eye Movements in the Reading of Chinese


Student thesis: Doctoral Thesis

View graph of relations



Awarding Institution
Award date17 May 2017


The typographic features of written Chinese texts significantly impact legibility and readability and essentially influence reading behaviours. Setting the typographic features within their optimal ranges in experiments on Chinese reading is critical to ensure the validity, reliability and consistency of experimental outcomes. Nevertheless, few previous reading experiments reported or included an examination of the readability under their typographic settings, leaving it unclear whether the obtained results would have any deviation from the true characteristics of normal reading.

The goal of this dissertation is to investigate the effects of four typographic features on eye movements in reading written Chinese: text-background luminance contrast (henceforth referred to as luminance contrast), word spacing, character size and character spacing. Specifically, the aim is to detect the optimal ranges of these typographic features for written Chinese. Moreover, readers’ eye movement characteristics are compared under optimal and non-optimal settings of these features to determine which aspects of reading are affected.

The first experiment examines the effect of luminance contrast on reading Chinese and finds (1) the optimal range of contrast ratio to be 2.454–7.523:1 and (2) low luminance contrast to significantly prolong fixation and overall reading time while leaving fixation frequency, fixation distribution and comprehension level almost unchanged. The cause to (2) slow visual information extraction due to low quality of reading stimulus at low luminance contrast. The result of the second experiment show that word spacing, regardless of the consistency of word segmentation, reduces reading efficiency compared to the normal unspaced layout, which is different from Bai et al.’s (2008) that the two conditions give similar reading efficiency. Compared to the normal unspaced condition, word spacing increases the reading speed on word, but slows down the overall reading, due to significantly more fixations on spaces. The comparison of three spaced conditions (i.e., character spaced, non-word spaced and word spaced) reveals that the ratio of spaces to characters modulates fixation distribution: the more spaces, the more fixations on spaces. The third experiment examines the effects of character size and character spacing, and finds the optimal range of character size to be a visual angle of 0.6-1° and the optimal character spacing to be 0. Reading proficiency decreases along with increase of character spacing, and an increase in either variable of these two narrows perceptual span and leads to increased number of fixations. Fixation distribution is also affected by these two variables, but in an opposite way: the larger the character size/the smaller the character spacing, the larger the proportion of fixations on characters. This appears to suggest that saccade planning in Chinese reading is largely affected by spatial factors.

In summary, the above outcomes help fill the gap between the research in human factors and psycholinguistics concerning Chinese reading, and suggest an optimal setting for each of these typographic features for our future reading experiments to increase their validity and reliability.