NotesFAQContact Us
Collection
Advanced
Search Tips
ERIC Number: ED409362
Record Type: Non-Journal
Publication Date: 1997-Mar
Pages: 21
Abstractor: N/A
Reference Count: N/A
ISBN: N/A
ISSN: N/A
Effects of Item and Response Set Reversals on Survey Statistics.
Barnette, J. Jackson
The controversy regarding reverse or negatively-worded survey stems has been around for several decades. The practice has been used to guard against acquiescent or response set behaviors. A 20-item, 5-point Likert item survey was designed and the stems and response sets were varied in a 2 by 3 design. One independent variable was type of item stem: one level had all direct-worded stems and the other had, randomly determined, half direct and half reverse-worded stems. The other independent variable was response set type. One level had all response sets going "strongly disagree" (SD) to "strongly agree" (SA), one had all response sets going SA to SD, and the third had, randomly determined, half going SD to SA and half going SA to SD. The surveys were administered to 687 subjects. The form each subject received was determined randomly. Responses were scored so that all were in agreement with the direct or positive form of the item stem. Item means were lower for the all direct-worded surveys compared with the half direct, half reverse-worded stems. The survey with the all direct stems and half SD-SA, half SA-SD response sets had the highest item variable. However, the most important finding was that the survey with the lowest reliability was the one with half direct and half-reverse worded stems with half SD-SA and half SA-SD response sets, while the survey with the highest reliability was the survey with all direct-worded stems with half SD-SA and half SA-SD response sets. This would indicate that the use of a combination of all direct-worded stems and half of the response sets going in one direction and half in the other direction may be a better way of guarding against acquiescence and response set behaviors. (Contains eight tables and nine references.) (Author/SLD)
Publication Type: Reports - Evaluative; Speeches/Meeting Papers
Education Level: N/A
Audience: N/A
Language: English
Sponsor: N/A
Authoring Institution: N/A