Latent Growth Modeling of Item Process Data Derived From Eye-tracking Technology: An Experimental Study Investigating Reading Behavior of Examinees When Answering A Multiple-Choice Test Item

This study illustrates how eye-tracking data can be translated to “item process data” for multiple-choice test items the link between subjects’ item responses and reading behavior. Several modes of analysis were used hypothesized added value using identify patterns subjects testing whether some variances in these could explained by being correct or incorrect. Application from a larger experiment included seventy-one two measuring comprehension ability English as second language. The collected processed one at time. For each item, first, subject’s recorded eye movements coded into set Area Interests (AOIs), segmenting lines stem individual choices. Next, fixation times on AOIs time stamped seconds, indicating when what order gaze had fixated AOI until choice was marked answer, ending encounter. Utilizing findings initial screening following descriptive analyses, nested Latent Growth Curve models fit Item 1 2 data. Model indices, random intercepts, slopes, residuals computed mean log across presentation AOIs. estimates describe shape change predictive power response. Albeit limited due two-item – experiement setting, show that utilizing item-level very useful defining common uncommon answer response is an important predictor associating behavior ability.