Determination of threshold of internal validity's rates for four-tier diagnostic chemistry test by response time

In the last twenty-five years, many measurement methods have been developed to identify misconceptions and these methods have been used in evaluation processes. Examples of these methods are interviews, open-ended exams, multiple choice tests, two-tier tests, three-tier tests, and four-tier tests. When participants encounter a test item, they respond to it in two ways: the first is the quick guessing behavior and the other is the solution behavior. Calculation of the threshold value plays a very important role in understanding the threshold value to understand whether the participant exhibits fast guessing or solution behaviour. A participant who responded to the test item in a time below the threshold calculated for the test items showed fast guessing behavior, and a participant who responded to the test item in a time above the threshold showed solution behavior. The threshold value is often used in multiple choice tests. In this study, the threshold value was used for the four-tier test. This study aimed to determine the threshold values of the internal validity rates of the four-tier chemistry diagnostic test and the multiple-choice chemistry test according to the Time to Response and Cox-Hazard Model. This study was conducted by descriptive survey method. Data collection tools were the 9-item Four-Tier Diagnostic Chemistry Test (FTDCT) and the Multiple Choice Chemistry Test (MCCT). The FTDCT was developed by Ünsal (2019). MCCT was adapted from the stepwise test and the first step of the FTDCT was used as test items. In this study, the scientific knowledge reliability coeficient of KR-20 for FTDCT was 0.460. Similarly, the misconception reliability of KR-20 for FTDCT was 0.570. Otherhand, the reliability coefficient of KR20 for the MCCT was calculated by 0.520. The study was conducted with the participation of pre-service science teachers studying at Dokuz Eylül University, Buca Faculty of Education in the 2020-2021 academic year. 75 pre-service teachers for the FTDCT and 74 pre-service teachers for the MCCT participated to tests. Conclusion, this study revealed that the threshold values of the internal validity rates of the FTDCT were 16 seconds at the scientific knowledge level, 17 seconds at the false positive level, and 28 seconds at the true negative level. The threshold value at the false negative level could not be determined. According to the MCCT, the response effort of the participants was calculated as 18 seconds. In this study, since the participants were not given the right to return to the test items, each participant had to answer the test items once and was not given the right to change their answer. Therefore, if each test item is given the right to return, there may be a change in the internal validity rates. In future studies, the scope of the study can be expanded according to the test application form.