Cheating detection in large-scale assessment received considerable attention in the extant literature. However, none of the previous studies in this line of research investigated the stacking ensemble machine learning algorithm for cheating detection. Furthermore, no study addressed the issue of class imbalance using resampling. This study explored the application of the stacking ensemble machine learning algorithm to analyze the item response, response time, and augmented data of test-takers to detect cheating behaviors. The performance of the stacking method was compared with that of two other ensemble methods (bagging and boosting) as well as six base non-ensemble machine learning algorithms. Issues related to class imbalance and input features were addressed. The study results indicated that stacking, resampling, and feature sets including augmented summary data generally performed better than its counterparts in cheating detection. Compared with other competing machine learning algorithms investigated in this study, the meta-model from stacking using discriminant analysis based on the top two base models—Gradient Boosting and Random Forest—generally performed the best when item responses and the augmented summary statistics were used as the input features with an under-sampling ratio of 10:1 among all the study conditions.

Editorial: Process Data in Educational and Psychological Measurement

This study explored the impact of partial credit scoring of one type of innovative items (multiple-response items) in a computerized adaptive version of a large-scale licensure pretest and operational test settings. The impacts of partial credit scoring on the estimation of the ability parameters and classification decisions in operational test settings were explored in one real data analysis and two simulation studies when two different polytomous scoring algorithms, automated polytomous scoring and rater-generated polytomous scoring, were applied. For the real data analyses, the ability estimates from dichotomous and polytomous scoring were highly correlated; the classification consistency between different scoring algorithms was nearly perfect. Information distribution changed slightly in the operational item bank. In the two simulation studies comparing each polytomous scoring with dichotomous scoring, the ability estimates resulting from polytomous scoring had slightly higher measurement precision than those resulting from dichotomous scoring. The practical impact related to classification decision was minor because of the extremely small number of items that could be scored polytomously in this current study.