Graphical modeling for item difficulty in medical faculty exams.

BACKGROUND: There are different indexes used in the evaluation of exam results. One important index is the difficulty level of the item that is also used in this study to obtain control charts. This article offers some suggestions for the improvement of multiple-choice tests using item analysis statistics. MATERIALS AND METHODS: The graphical modeling is important for the rapid and comparative evaluation of test results. The control chart is a tool that can be used to sharpen our teaching and testing skills by inspecting the weaknesses of measurements and producing reliable items. The research data for the application of control charts were obtained using the results of the fourth and fifth-grade student's exams at Ondokuz Mayis University, Faculty of Medicine. I-chart or moving range chart (MR) is preferred for whole variable data. RESULTS: It is seen that all observations are within control limits for I-chart, but three points on MR-chart are settled on the LCL. Using X--chart with subgroups, it was determined that control measurements were within the upper and lower limits in both charts. The difficulty levels of items were examined by obtaining different variable control charts. The difficulty level of the two items exceeded the upper control limit in R- and S-charts. CONCLUSION: The control charts have the advantage for classifying items as acceptable or unacceptable based on item difficulty criteria.

Mixed multivariate generalized linear models for assessing lower-limb arterial stenoses.

Experiments and observational studies often involve gathering information on several response variables, enabling us to model their dependence on observable predictor variables. Despite associations between the response variables, they are often analysed separately using general and generalized linear models. This paper investigates applications of multivariate regression analysis to improve the accuracy of predictions and decisions, in the specific context of diagnosing arterial stenoses in human legs. Two basic models are developed for this application, using (i) four binary responses and (ii) a mixture of two binary and two normal responses. The results clearly demonstrate the potential advantages offered by this approach.