NotesFAQContact Us
Search Tips
ERIC Number: ED218361
Record Type: Non-Journal
Publication Date: 1982-Mar
Pages: 18
Abstractor: N/A
Reference Count: N/A
A Study of the Reliability of Nedelsky's Method for Choosing a Passing Score.
Livingston, Samuel A.; Kastrinos, William
Leo Nedelsky developed a method for determining absolute grading standards for multiple choice tests. His method required a group of judges to examine each test question and eliminate those responses which the lowest D- student should be able to reject as incorrect. The correct answer probabilities remaining were used in computing an expected test score for the hypothetical test taker. The passing score was chosen to give the "F-D student" some probability of passing the test. Nedelsky's method of choosing passing scores are being used by many licensing and certification boards. Alternate forms of a test, containing 25 ittms in common, were used to measure the reliability of passing scores chosen by this method. Experts judged the tests. There were large and consistent differences between the judges for their passing scores. The experts tended to set higher passing scores the second time they judged the questions and to leave different wrong answer choices unmarked. The results of this study indicated users of Nedelsky's method should consider a two stage judgment procedure. (Author/DWH)
Publication Type: Reports - Research
Education Level: N/A
Audience: N/A
Language: English
Sponsor: N/A
Authoring Institution: Educational Testing Service, Princeton, NJ. Center for Occupational and Professional Assessment.