NotesFAQContact Us
Search Tips
ERIC Number: ED402321
Record Type: Non-Journal
Publication Date: 1996-Apr
Pages: 48
Abstractor: N/A
Reference Count: N/A
Factor Analytic Methods for Determining Whether Multiple-Choice and Constructed-Response Tests Measure the Same Construct.
Manhart, Jim J.
The relationship between multiple-choice and constructed-response science tests was investigated using confirmatory factor analysis. The tests were given to 872 students in grades 9 through 12. Each test was divided into several parcels of items. The fit of a one-factor model (parcels of both tests loading on the same factor) was compared with the fit of a two-factor model (parcels of each test loading on a different factor). Inspection of chi-square data and standardized residuals led to the conclusion that the two-factor model was generally more appropriate for explaining the covariance between the parcels than the one-factor model. The conclusion about the number of factors was consistent across six methods that varied the number of parcels used and how items were assigned to the parcels. Two appendixes present the test specifications for the multiple-choice and constructed-response tests. (Contains 8 tables and 16 references.) (Author/SLD)
Publication Type: Reports - Evaluative; Speeches/Meeting Papers
Education Level: N/A
Audience: N/A
Language: English
Sponsor: N/A
Authoring Institution: N/A