论文信息 - Study on Sample Size of Candidates Oriented to Online Test

Study on Sample Size of Candidates Oriented to Online Test

The sample size directly affects the accuracy of the item parameter estimation of the online test. We simulate several tests comprised of 5 items, 9 items, 21 items and 41 items with different difficulties. Then, the Monte Carlo method is used to simulate candidates’ answers to test items by Rasch response model. From this, we calculate the error metric of the test item distribution under different sample sizes. By comparing the real data with simulation data, we find that the fitting effect is good.

Xiaoxia Zhao | Jimei Li | Yong Qiao | Yanbin Feng

[1] Alper Sahin,et al. The Effects of Test Length and Sample Size on Item Parameters in Item Response Theory , 2016 .

[2] Anna Brown,et al. Handbook of Item Response Theory Modeling : Applications to Typical Performance Assessment , 2014 .

[3] Steven P. Reise and Dennis A. Revicki: Handbook of item response theory modeling: applications to typical performance assessment. Multivariate applications series , 2015, Quality of Life Research.

[4] Ping Guo,et al. The research and application of online examination and monitoring system , 2008, 2008 IEEE International Symposium on IT in Medicine and Education.

[5] David Andrich,et al. Rasch Models For Measurement , 1988 .

[6] Randall E. Schumacker,et al. Ability Estimation Under Different Item Parameterization and Scoring Models , 2004 .

[7] Liu Xinxin,et al. Algorithm and strategy research of generating test paper for large scale online examination , 2011, Proceedings of 2011 International Conference on Electronic & Mechanical Engineering and Information Technology.

[8] Fritz Drasgow,et al. Recovery of Two- and Three-Parameter Logistic Item Characteristic Curves: A Monte Carlo Study , 1982 .