top of page

Click to see the full paper.

KMV18 heatmap.jpg

Click to see the full paper.

Figure 3. Heat map of the relative importance of text- and participant-level predictors of word processing effort across the eye-movement record. Separate models were fit to the subset of the data containing long words (more than 6 character long) and short words.



Back in 2009, when we began this work, computational models of eye-movement control focused entirely on properties of the text (especially word frequency and length) to predict eye-movement behavior.  Ten years later, this is still largely true.  Yet there is now ample evidence that person-level variables play a large role in oculomotor programming.  We have shown that abilities in Rapid Letter Naming (a.k.a. RAN) and single-word reading account for substantially more variance in the earliest eye-movement measures than text-variables such as word length and word frequency.  Even in later measures, the effects of these skills are second only to word length.  Moreover, in Kuperman and Van Dyke (2011), these skills had a tremendous influence on interactions with text-level variables, often varying the effect size of word length by a factor of 3 and of word frequency by a factor of 4.


We verified these results using an alternative analysis method (Random Forests), which allows consideration of many predictors while controlling for model overfitting and collinearity of predictors. As before, word reading (10th row) and RAN (rows 14-17) were ranked as highly important predictors of early eye-movement measures.


We further explored the cause of the RAN-reading relationship and found evidence that oculomotor control accounts for ~45% of the variance accounted for in the RAN task, suggesting that the individual ability to coordinate rapid sequential eye movements is a crucial component in fluent reading.

Current Collaborators

Representative Publications

Schmidtke, D., Van Dyke, J. A.,& Kuperman, V. (2019, July 16). CompLex: An eye-movement database of compound word reading in English. [osf]

Kuperman, V., Matsuki, K., & Van Dyke, J.A (2018). Contributions of reader- and text-level characteristics to eye-movement patterns during passage reading. Journal of Experimental Psychology: Learning, Memory, and Cognition.  [publisher | pubmed]

Henry, R., Van Dyke, J. A., & Kuperman, V. (2018). Oculomotor planning in RAN and reading: A strong test of the visual scanning hypothesis. Reading and Writing, 31(7), 1619-1643. [publisher| pubmed}

Kuperman, V., Van Dyke, J.A.,Henry, R. (2016). Eye-movement control in RAN and Reading. Scientific Studies of Reading, 20(2), 173-188. [publisher | pubmed]

Kuperman, V., & Van Dyke, J.A.(2013). Reassessing word frequency as a determinant of word recognition for skilled and unskilled readers. Journal of Experimental Psychology: Human Performance and Perception, 39(3), 802-823. [publisher | pubmed]

Kuperman, V., & Van Dyke, J.A.(2011). Effects of individual differences in verbal skills on eye-movement patterns during sentence reading. Journal of Memory and Language, 65(1), 42-73. [publisher | pubmed]

bottom of page