In computerized adaptive testing (CAT), aberrant responses such as careless errors and lucky guesses may cause significant ability estimation biases in the dynamic administration of test items. We investigated the robustness of the 4-parameter logistic item response theory (4PL IRT; Barton & Lord, 1981) model in comparison with the 3-parameter logistic (3PL) IRT model (Birnbaum, 1968). We applied additional precision and efficiency measures to evaluate the 4PL IRT model. We measured the precision of CAT with respect to the estimation bias and mean absolute differences (MAD) between estimated and actual abilities. An improvement in administrative efficiency is reflected in fewer items being required for satisfying the stopping rule. Our results indicate that the 4PL IRT model provides a more efficient and robust ability estimation than the 3PL model.

No abstract

No abstract

In this study we investigated whether guided collaboration would enhance elementary school students' achievement in learning to program in MSWLogo. Nintyfour sixth graders were randomly assigned to the individual-learning group, the free-collaboration group or the guided-collaboration group. All students were taught to follow the problem-solving steps of analysis, design, coding, debugging, and reflection, but only students of the guided-collaboration group were provided a worksheet that aimed to force students to carry out all specified activities at each step. A statistical analysis of students' test scores showed that the guided-collaboration group significantly outperformed the other two groups. The differences among the low achievers of the three groups were especially significant, indicating that they had benefited the most from guided collaboration. It was also found that guided collaboration promoted meaningful discussions among students, which had contributed to their better understanding of programming concepts and problem-solving skills, and hence better test scores.