ERIC Number: ED193263
Record Type: RIE
Publication Date: 1980
Reference Count: 0
Student Ratings of College Courses and College Teachers: Reliability, Generalizability, Validity and Utility.
Wood, Peter H.; And Others
A review of the research literature considers the following questions: (1) what is the general quality of the information resulting from systematic student evaluation of faculty effectiveness?; and (2) what are the consequences of soliciting this information and making it available to those people who may utilize it in their decision making procedures? The internal consistency of most ratings scales, and many subscales, often exceed .9, and, when lower, often can be increased by increasing the number of ratings items in a particular scale. However, this indicator of reliability produces a considerable overestimation of ratings stability when independent measures of teachers and/or courses are desired. Generalizability theory has begun to be employed to supplement and to extend the more traditional approaches to reliability. Factors such as expected grades, class-size, and the students' pre-enrollment attitudes are likely to influence student ratings. Much of the instability and potential invalidity may be reduced through the standardization of administrative procedures and the use of global rating items for normative or personnel purposes. There appears, however, to be little evidence that student ratings are weighted very heavily in the decision-making processes of personnel committees at the college level. (RL)
Publication Type: Information Analyses
Education Level: N/A
Authoring Institution: Bowling Green State Univ., OH. Coll. of Education.