NotesFAQContact Us
Collection
Advanced
Search Tips
ERIC Number: ED382687
Record Type: RIE
Publication Date: 1995-Apr
Pages: 25
Abstractor: N/A
Reference Count: N/A
ISBN: N/A
ISSN: N/A
Performance of Angoff Model IV Linear Test Equating Using Total Test and Content Dimensional Sub-Test Designs in Small Groups of Examinees.
Motika, Robert T.; Chason, Walter M.
Test data from 200 examinees from the Spanish Teacher Certification Examination and 75 examinees from the French Teacher Certification Examination were used in a study of scale drift in sequentially equated test forms. Using sampling with replacement, 1,000 samples of 100 examinees each for Spanish and 1,000 samples of 50 each for French were created, and each sample was equated using Form A as the base form and equating through form B and form C back to form A. The Spanish forms were then broken down by item into three validated domains while the French forms were categorized into two domains. The forms were then equated by domain and the resulting equated domain scores summed to provide an estimate of total test performance. It was found that the whole test equating provided a better estimate based on scale drift than the summed scores of the equated domains. Nine tables. (Contains six references.) (Author)
Publication Type: Reports - Research; Speeches/Meeting Papers
Education Level: N/A
Audience: N/A
Language: English
Sponsor: N/A
Authoring Institution: N/A
Identifiers: Angoff Methods; Linear Equating Method; Scale Drift
Note: Paper presented at the Annual Meeting of the American Educational Research Association (San Francisco, CA, April 18-22, 1995).