Effects on response time and accuracy of technology-enhanced cloze tests: an eye-tracking study.

The transition from paper-based tests to corresponding computer-administered tests allows for the incorporation of improved interfaces that support response making. The main research question is whether innovative interfaces affect test response time and/or response accuracy. This study compared performance on banked cloze tests using a conventional interface (based on paper-based formats for responding by writing a number in a box) versus cloze tests with improved interfaces (based on computer-based affordances such as dragging and dropping responses to fill in a blank). In a banked cloze test, the left side of the page shows a text with words deleted and replaced with blanks that are numbered, and the right side shows the word list with a space to write in the corresponding number. In Experiment 1, 56 fourth graders in the conventional group responded more slowly but just as accurately, and spent more time looking at the word list on the right of the screen but spent equivalent time looking at the text as compared to a group that took the test with an improved interface. In Experiment 2, the same pattern of results was replicated with 148 sixth graders and for each of three versions of improved interfaces as compared to the conventional interface. Results support the idea that the improved interface affected the response execution phase but not the response development phase of performance on the cloze test. [ABSTRACT FROM AUTHOR]