Reports a study to explore student's computational performance on computer-administered testing situations while at the computer. Students from grades 3-6 (n=100) were randomly assigned to treatment order, computer assessment first or paper/pencil assessment first. Significant differences occurred in three areas: completion time, number of mental…

Describes responses of fifth to eighth grade students to different types of test items requiring estimation. Reports that performance differed by item format, types of numbers and operations in the items, and grade level of students. (Author/YP)