| Summary: | The aim of this research is examine the correlation between student achievements and the items that construct the test according to the Classical Test Theory and Item Response Theory. The survey method has been chosen for this research. Study group includes 231students which studying in Eskişehir Osmangazi University and attending to Measurement and Evaluation class. A 40 multiple-choice items test has been developed and administrated. The test aims to measure the academic achievement on measurement and evaluation lesson. Scoring has been made according to Classical Test Theory and Item Response Theory, and then scores have been compared. Analyses show that the correlation between the scores is positive, very high and significant. After the item analyses, 5 items have been opt out of the test by classical test theory. And 14 items have been opt out of the test by item response theory. After opting out, scoring has been made again and the correlation between the scores is positive, very high and significant again. When we consider the test scores to pass – fail decision; two theories only differ 12.6% of students and the same decision has been made with 87.4% of students.
|
|---|
