ERIC Number: ED090246
Record Type: Non-Journal
Publication Date: 1973
Reference Count: N/A
The Use of Factor Analysis to Salvage Poor Guttman Scales: Can it Really Work.
Kayser, Brian D.
The Guttman model of scale analysis has found continued use in sociological analysis despite criticisms placed against it. An empirical example is provided of the use of factor analysis with Guttman scaling even taking into account the criticisms of both very restricted item number of dichotomous responses. Data came from a questionnaire using three theoretical law scales. Attitudes, values, and aspirations regarding the legal institution were solicited from high school students. Students indicated on the questionnaire if they strongly agreed, agreed, were undecided, disagreed, or strongly disagreed with each item. The responses were coded such that the higher the score, the more favorable the attitude toward the legal institution. Factor analysis of all items using principal axis with varimax rotation was used. Scale analysis revealed that the scales were not Guttman. Although the scale and item reproducibilities were high, the scale scalabilities were not significantly greater than 0.60. Delineation of the structure of the scale is given as a justification for use of factor analysis with Guttman scales. It appears that the use of factor analysis with Guttman scales is not a desirable technique of scaling. A better approach would be the single use of factor analysis in order to help construct single factor scales of greater length. (Author/RC)
Publication Type: N/A
Education Level: N/A
Sponsor: Economic Development Administration (DOC), Washington, DC. Office of Economic Research.; National Inst. of Mental Health (DHEW), Bethesda, MD.
Authoring Institution: Wisconsin Univ., Madison. Center of Applied Sociology.