The purpose of this study was to investigate the effects of test length on the estimation of functional reading ability levels and mastery/nonmastery classifications of ninth grade students in the state of Maryland with short-form tests. developed by using the oneparameter item response model.
Using the i tern responses of approximately 36,000 students to 75 items on a functional reading test, item parameters were estimated with the one-parameter i tern response model. Three nonoverlapping short-form tests of 10, 20, and 30 i terns were developed using i terns targeted at the cut-score of the test. The study investigated the extent to which estimates of pupil functional reading ability levels and mastery/nonmastery classifications obtained from three short-form tests were the same as or related to those obtained on the original 75-item test. Three nonoverlapping samples of 5,000 students each were used to make the comparisons. The extent to which estimates of pupil performance on the short-form measures were the same as that on the original tests was analyzed using a m ul ti-variate analysis of variance design. The results showed that the ability estimates obtained on each of the short-form tests differed significantly (p(.000) from that obtained on the original test. The differences were, however, trivial, measuring less than .06 of the standard deviation of the shorter test.
Pearson's product moment correlation coefficients obtained in this study were, on average, .80, .89, and .94 between the original test (75 items) and the 10-, 20-, and 30-item tests, respectively.
Analysis of the mastery/nonmastery classifications resulted in observed indices of agreement be tween the short- and long-form tests ranging from .82 for the 10-item test to .91 for the 30-item test. Kappa indices of agreement between the the short- and long-form measures ranged from .64 for the 10-item test to .81 for the 30-item test.
The study concluded that there is a relationship be tween test length and estimation of pupil functional reading ability and student mastery/nonmastery classifications. It is proposed, however, that a substantial reduction in testing time and student testing burden can be realized by using short-form tests developed and administered in a manner described in the study. / Doctor of Education
| Identifer | oai:union.ndltd.org:VTETD/oai:vtechworks.lib.vt.edu:10919/88651 |
| Date | January 1984 |
| Creators | Tompkins, Leroy J. |
| Contributors | Curriculum and Instruction, Hutson, Barbara A., Gatewood, Thomas E., Williams, Paul L., McKeen, Ronald L., Fortune, Jimmie C. |
| Publisher | Virginia Polytechnic Institute and State University |
| Source Sets | Virginia Tech Theses and Dissertation |
| Language | en_US |
| Detected Language | English |
| Type | Dissertation, Text |
| Format | ix, 135 leaves, application/pdf, application/pdf |
| Rights | In Copyright, http://rightsstatements.org/vocab/InC/1.0/ |
| Relation | OCLC# 11298662 |
Page generated in 0.0018 seconds