EROĞLU, Melek Gülşah Gazi Üniversitesi - Gazi Eğitim Fakültesi - E B B Eğitimde Ölçme ve Değerlendirme A B D, Turkey , KELECİOĞLU, Hülya Hacettepe Üniversitesi - Eğitim Fakültesi - E BB Eğitimde Ölçme ve Değerlendirme A B D, Turkey

Comparison Of Different Test Termination Rules in Terms of Measurement Precision and Test Length in Computerized Adaptive Testing

30504

In adaptive testing, there exists a test algorithm different than the classical tests. The test algorithm consists of three parts which are starting, resuming and termination. The aim of this study is to measure the effect of different termination rules on measurement precision and test length in computer adaptive testing. The research was implemented as a simulation study. Five different termination rules have been used for the study which are: fixed length, standard error, standard error-least item, theta convergence and theta convergence-least item. Different conditions are in place in each termination rule and a total of 12 conditions are compared. Additionally, in comparing termination rules, different item pools (250 and 500) and ability estimation methods (Maksimum Likelihood Estimation and Expected a Posteriori) have been selected since these are critical in the algorithms of Computer Adaptive Testing. RMSE, bias and fidelity values were calculated for the measurement precision and test lengths were obtained and compared for each of the CAT implementation. As a result, for the 20 item fixed length, 0,220 standard error and 0,02 theta convergence termination conditions RMSE and bias values are small but fidelity factors are not significantly affected. And with the addition of the least item factor, better results were achieved in some of the termination conditions in terms of measurement precision. The test length is observed to be negatively correlated with the RMSE values. In the same termination conditions, with the increases in item pool generally smaller RMSE and bias values were for measurement precision were achieved. Not a significant change was observed in the evaluation of the effect of the starting rules. The preference of Expected a Posteriori method for the ability estimation is observed to cause a drop in values for RMSE and bias values.

31

Computer Adaptive Testing (CAT) , Termination Rules , Measurement Precision , Test Length

Journal Of Uludag University Faculty Of Education

52