Measurement Approach to the Comparisons of Career Anchor Models

• Main Author: Cai, Mengfei
• Format: Others
• Published: BYU ScholarsArchive 2012
• Subjects: career anchor; measurement methods; model fit; age differences; Psychology
• Online Access: https://scholarsarchive.byu.edu/etd/3581; https://scholarsarchive.byu.edu/cgi/viewcontent.cgi?article=4580&context=etd

The career anchors concept is an approach to understanding career orientation and motivation. The original career anchor model was introduced by Schein in 1974. Several investigators have created revisions of the model to make it more useful. This dissertation is a continuation of the quest to evaluate the original model and the revised models with respect to empirical support. This study is the first of two studies in which measurement methods are devised to solve the psychometric problems of previous measures. In this study we create and test an "economic exchange" model to correct the problem of acquiescent bias. We test five career anchor models and this new scaling method against two sets of data. The first consists of data from 330 participants we collect in the present study, and the other is a set of correlation matrices from Barclay's dissertation meta-analysis of six previous studies from the literature. We find that the economic exchange method creates greater variances in the ratings (both within each person and across persons) as predicted, but the hypothesis of predicted increase in the range of correlation coefficients for this method is not supported. In its present form the economic exchange method is not found to be superior to the standard Likert scale method. In addition, the oppositionality of career anchor choices does not increase for older respondents as expected. From a confirmatory factor analysis test of goodness of fit of the five models against the six datasets of this dissertation and the six studies from Barclay's meta-analysis, we find no evidence for one best career anchors model. That is, the five competing theoretical models seem to each be "best" in some situations or populations.