NotesFAQContact Us
Search Tips
ERIC Number: ED430046
Record Type: Non-Journal
Publication Date: 1999-Apr
Pages: 32
Abstractor: N/A
Reference Count: N/A
Evaluating College Services Using Student Ratings with Incomplete Data: An IRT Rating Scale Model Approach.
Sun, Anji; Schulz, E. Matthew
Missing data create problems for the interpretations and inferences of survey results, especially if the amount of missing data is substantial. This study considers issues of missing data when comparisons are made across survey items. It is suggested that if the groups responding to different items differ in their tendency to use one end of the measurement scale or the other, named as "pleasability" according to the specific nature of the data used in the study, then the comparisons of these items should be adjusted for this difference. Data for the study were obtained from the Student Opinion Survey processing history files of ACT Inc. for 10 institutions. Students at these colleges had responded to the survey about college services used, but the average number of students responding to all items was less than half the total number of respondents. The results of the study show a consistent difference in pleasability among the respondents and that an item response theory rating scale model can improve comparisons among survey items when the data consist of ratings on a Likert scale and respondents rate only selected items. (Contains 4 figures, 11 tables, and 17 references.) (Author/SLD)
Publication Type: Reports - Research; Speeches/Meeting Papers
Education Level: N/A
Audience: N/A
Language: English
Sponsor: N/A
Authoring Institution: N/A