NotesFAQContact Us
Collection
Advanced
Search Tips
ERIC Number: ED504912
Record Type: Non-Journal
Publication Date: 2009-Feb
Pages: 18
Abstractor: As Provided
Reference Count: 53
ISBN: N/A
ISSN: N/A
The Gulliver Effect: The Impact of Error in an Elephantine Subpopulation on Estimates for Lilliputian Subpopulations
Micceri, Theodore; Parasher, Pradnya; Waugh, Gordon W.; Herreid, Charlene
Online Submission, Paper presented at the Annual Meeting of the Florida Association for Institutional Research (Cocoa Beach, FL, Feb 25-27, 2009)
An extensive review of the research literature and a study comparing over 36,000 survey responses with archival true scores indicated that one should expect a minimum of at least three percent random error for the least ambiguous of self-report measures. The Gulliver Effect occurs when a small proportion of error in a sizable subpopulation exerts a large influence on one or more Lilliputian/small subpopulations. As a simple example, if three percent random measurement error occurs for a sample of 1,000 on a race/ethnicity item having two groups (white and minority), where the majority group make up 90 percent of the total population, then 27 majority subjects (three percent) will be erroneously classified as minorities and three minorities (three percent) will be incorrectly classified as majority. Assuming no missing data, the study will therefore report 876 majority respondents (873 majority + 3 incorrectly classified minority respondents) and 124 minority respondents (97 minority + 27 incorrectly classified majority). Although the incorrect majority respondents percentage is small (3/873 = 0.3%), the erroneous minority classification is substantial (27/97 = 27.8%). Such a large proportion of incorrectly classified respondents may alter estimates of measured differences between groups. The Gulliver Effect may occur whenever small subpopulations are of interest, whether this involves endangered species, rare diseases or unusual crimes such as the kidnapping of juveniles. (Contains 2 endnotes and 1 footnote.)
Publication Type: Reports - Evaluative; Speeches/Meeting Papers
Education Level: N/A
Audience: N/A
Language: English
Sponsor: N/A
Authoring Institution: N/A