NotesFAQContact Us
Search Tips
ERIC Number: ED213296
Record Type: RIE
Publication Date: 1981-Nov
Pages: 7
Abstractor: N/A
Reference Count: 0
Reliability, Validity, and Generalizability of Student Ratings of Instruction. IDEA Paper No. 6.
Aubrecht, Judith D.
The twin issues of reliability (consistency) and validity (relevance) of student ratings of teacher performance, and the overall issues of generalizability, continue to arise in assessing the usefulness of this faculty evaluation technique. Recent literature addressed four basic questions regarding these issues as they pertain to two faculty evaluation matters, personnel decisions and instructional improvement. The four questions are: (1) Are student ratings biased? (2) Do they agree with the ratings of other groups? (3) Do different kinds of student ratings yield similar information? and (4) Do these ratings help teachers improve instruction? It is neither easy nor very productive to separate the concepts of reliability and validity, and classical reliability theory has evolved into a theory of generalizability, which looks at teacher effect, course effect, student effect, and/or item effect in instruction. Research strengthens the idea that student ratings are both consistent and reliable in teacher evaluation. By and large, strong biasing variables can be controlled statistically, and there is considerable agreement both between student ratings and other groups, and between types of student rating. It also appears that student ratings have served to improve instruction somewhat. More research on generalizability is recommended and appears likely. (MSE)
Center for Faculty Evaluation and Development, Kansas State University, 1623 Anderson Ave., Manhattan, KS 66502-4098 ($.25 each, minimum order 25 copies).
Publication Type: Information Analyses
Education Level: N/A
Audience: N/A
Language: English
Sponsor: N/A
Authoring Institution: Kansas State Univ., Manhattan. Center for Faculty Evaluation and Development in Higher Education.