English Deutsch Français Italiano Español Português 繁體中文 Bahasa Indonesia Tiếng Việt ภาษาไทย
All categories

2006-12-03 18:40:54 · 2 answers · asked by Anonymous in Education & Reference Standards & Testing

2 answers

Questions about a particular subject
Answers relevant to the questions
A defined percentage of correct answers to pass

2006-12-03 18:48:21 · answer #1 · answered by Boring Old Fart 3 · 0 0

A professional test developer would ask about (1) the validity of the test and (2) its reliability, expressed as the standard error of measurement (SEM). These are the key characteristics.

Assuming the validity is sufficiently high and the SEM is low, the next question is about the results of the item analysis. Here you want to know (1) the difficulty of each item, (2) the discrimination index of each item, and (3) the effectiveness of the distractors if multiple-choice items are used.

There are many additional "key" factors once the test has met the basic requirements (above) for a good test. Most of these deal with the administration of the test. For example, in an employment setting, if the test has disparate impact on a protected group, it may have to be withdrawn in favor of another test that does not. I hope this answers your question. Best wishes.

If you want to know more, go to http://www.siop.org

2006-12-06 14:29:14 · answer #2 · answered by Robert W 1 · 0 0

fedest.com, questions and answers