Eye Movements of Developing Chinese Readers: Effects of Word Frequency and Predictability

ABSTRACT

The frequency and contextual predictability of words have a fundamental role in determining where and when the eyes move during reading in both alphabetic and non-alphabetic languages. However, surprising little is known about the how the influence of these variables develops, although this is important for understanding how children learn to read. Accordingly, to gain insight into their use during reading development, we examined the effects of orthogonally manipulating the frequency and contextual predictability of a specific target word in sentences on the eye movements of developing Chinese readers. The findings show that both factors influence eye movement behavior associated with the early processing of words during reading, but that effects of contextual predictability are mediated by the lexical frequency of words. We consider these effects in the context of visual and linguistic demands associated with reading Chinese and in relation to current models of eye movement control during reading.

Acknowledgments

The research was supported by a Chinese Ministry of Education Project of Key Research Institute of Humanities and Social Sciences in Universities (grant number 15JJD190003) Guoli Yan, and 1000 Talents Visiting Professorship to Kevin Paterson. We also thank Sascha Schroeder and two anonymous reviewers for the insightful comments on earlier drafts of this article.

Data files and related resources are available from the University of Leicester online Figshare repository: https://leicester.figshare.com/s/979dd64534e3350ef379

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

1. Participants did not see an equal number of target words in each condition. We therefore repeated the data analyses reported here with a balanced dataset created by removing the extra word pair that unmatched the number of items per condition. The same pattern of effects was found, therefore, we decided to keep the original dataset.

2. Models with a maximum random structure were used for all reading times measures (SFD, GD, RPD, TRT). However, the random structure of models was trimmed to (1 + frequency × predictability | pp) + (1 + frequency + predictability |stim) for first-fixation duration; and trimmed to (1 + frequency | pp) + (1|stim) for measures of word-skipping, regressions-in and regressions-out.

Additional information

Funding

This work was supported by the Chinese Ministry of Education Project of Key Research Institute of Humanities and Social Sciences in Universities [15JJD190003]; 1000 Talents Visiting Professorship [1000 Talents Visiting Professorship].