Files in this item



application/pdf8018196.pdf (4MB)Restricted to U of Illinois
(no description provided)PDF


Title:Measuring Appropriateness in the Assessment of Job Satisfaction
Author(s):Parsons, Charles Kramer
Department / Program:Psychology
Degree Granting Institution:University of Illinois at Urbana-Champaign
Subject(s):Psychology, Industrial
Abstract:Latent trait theory or item characteristic curve theory has provided new insights into old problems in the measurement of mental aptitude. There are, however, several assumptions of the theory that may, on the surface, appear to restrict applications of the model. The assumptions that the test is unidimensional and that item responses are locally independent would seem to preclude the application of the theory from multidimensional job satisfaction measures. In direct opposition to this conservative view, this dissertation is written from the perspective that violations of these assumptions are relevant to statements about the robustness of the model rather than conclusions about its limited application. In the current study, the model is applied to the measurement of job satisfaction in an attempt to identify invalid response records.
Standard self-description inventories are problably the most frequently used indicators of job satisfaction. Their popularity is due to easy administration and the assumed comparability of scores across individuals and groups. With this approach, the scoring procedure is fixed across all research participants. There is no consideration that the same responses could have different meanings for different individuals. This neglect certainly affects the overall quality of the job satisfaction assessments.
There is some evidence that a group of statistics derivable from latent trait theory postulates, called appropriateness indices, can aid in the identification of invalid answer sheets to multiple choice aptitude tests. The logic would also seem to apply to identifying invalid answer sheets from the assessment of job satisfaction. The goal of this dissertation research is to examine the generalizability of earlier research on aptitude tests to job satisfaction measures and examine the utility of an appropriateness index as a data quality indicator for identifying errors in the prediction of employee turnover.
Data for this study were made available from a series of job attitude studies under the direction of Professor Charles Hulin, University of Illinois. Responses to 60 items of the Job Descriptive Index served as an index of job satisfaction. A sample of 1906 response records was selected for item parameter estimation. These records represented responses from both military and civilian personnel. Latent trait item parameters for the Job Descriptive Index were estimated using the maximum likelihood algorithm, LOGIST, available from Lord of Educational Testing Service. These parameter estimates were used to compute an appropriateness index (the geometric mean likelihood of a vector of Job Descriptive Index item responses) for each response record.
Several analyses were then conducted to investigate the properties of the appropriateness statistic in the job satisfaction data. Computer simulation provided the experimental control necessary to conclude that lower appropriateness indices were associated with response records that had been arbitrarily altered by the experimenter. These results were consistent with previous simulation results using mental aptitude tests. The appropriateness index also was related to race and education level of respondents. Blacks had lower indices than Whites, and lower education level was associated with lower appropriateness. A final study showed that there was little or no relation between appropriateness indices and errors of prediction.
Though conclusions are few at this point, this empirical investigation did suggest that further study of job satisfaction measures from the latent trait approach would be fruitful.
Issue Date:1980
Description:115 p.
Thesis (Ph.D.)--University of Illinois at Urbana-Champaign, 1980.
Other Identifier(s):(UMI)AAI8018196
Date Available in IDEALS:2014-12-13
Date Deposited:1980

This item appears in the following Collection(s)

Item Statistics