NotesFAQContact Us
Search Tips
ERIC Number: ED118620
Record Type: Non-Journal
Publication Date: 1974-Oct
Pages: 4
Abstractor: N/A
Reference Count: N/A
The Effect of Format Change on Test Item Difficulty.
Boyd, Joseph L., Jr.
This study evolved from a practical field situation that dictated an item format change. The item type in question is a variety of the multiple true-false item, widely used in state and municipal civil service examinations. Items are produced, not by combining pairs of independent true-false items, but as a means of salvaging not quite adequate four-choice multiple items. The original items may have no right answer: two, three, or four right answers; or one or two ambiguous or nonplausible responses. Items of this type were developed for use in a national testing program for automotive mechanics. After review by a number of test specialists and mechanics, it was decided that many items were faulty. A format variation was developed which seemed to be more clear cut, informal, and easier to read and understand. The revised items were used in a test battery. In practical terms, the overall effect of changing item format was to make the test items easier by an amount that would make mean percent correct scores higher by less than one percent. The real effect was to eliminate protests about the test questions. (RC)
Publication Type: Reports - Research
Education Level: N/A
Audience: N/A
Language: N/A
Sponsor: N/A
Authoring Institution: N/A