NotesFAQContact Us
Search Tips
ERIC Number: ED249286
Record Type: Non-Journal
Publication Date: 1984-Aug-27
Pages: 13
Abstractor: N/A
Reference Count: N/A
Computer-Based Test Interpretation and the Public Interest.
Mitchell, James V., Jr.
Computer-based test interpretation (CBTI) is discussed in terms of its potential dangers to the public interest, problems with professional review of CBTI systems, and needed policies for these systems. Several problems with CBTI systems are outlined: (1) they may be nicely packaged, but it is difficult to establish their value; (2) they do not allow the in-depth analysis of response data that a psychologist might require; (3) decision rule data are not available from the publishers; (4) the strength and statistical significance of the relationships on which these decision rules are based are often not clear or are questionable; (5) if the decision rules are modeled after the decision processes of an expert clinician, they are even more hidden than in an actuarial system; and (6) the validity of the interpretive report cannot be established. Because of the difficulties with reviewing CBTI systems, potential users cannot be guided by the advice of professionals. Policies should be developed by the profession regarding the disclosure of information about CBTI systems and the validity of their interpretive reports. (BW)
Publication Type: Speeches/Meeting Papers; Opinion Papers
Education Level: N/A
Audience: N/A
Language: English
Sponsor: N/A
Authoring Institution: N/A