Elites and Change in the Kentucky Mountains by Plunkett H. Dudley;Bowman Mary Jean; & Mary Jean Bowman

Author:Plunkett, H. Dudley;Bowman, Mary Jean; & Mary Jean Bowman [Plunkett, Dudley H. & Bowman, Mary Jean]
Language: eng
Format: epub
Publisher: University Press of Kentucky
Published: 2021-08-15T00:00:00+00:00

THE COMPONENTS ANALYSIS &

ATTITUDE CLUSTERS

A discussion of responses on particular attitude items grouped informally, as in Chapter 4, by a priori classifications provided a useful if impressionistic overview of how the various mountain elites view themselves and others, and of how they look at mountain prospects and some programs related thereto. However, it is all too easy to misinterpret responses on single items, projecting meanings that were not in respondents’ thinking, or assuming associations in attitudes that are not prevalent in fact. Components analysis of the inventory of attitudes provides a check on the validity of interpretations of particular measures, and it serves to identify dimensions (or clusters) of attitudes that may not have been anticipated but that challenge us to fresh thinking.

All those opinion items in the questionnaires and the interviews for mountain respondents that were common to the teachers and other occupational groups (except manufacturers) were subjected to a components analysis with varimax rotations. Eight factors were obtained, all of which appeared to be significant. Table 35 lists the factors, giving them interpretive titles and recording the highest-loading items (with their loadings) on each factor.1 It shows also the distributions of responses on those items, including proportions N.R.; the percentages total to 100 in each row. To allow easy cross-reference, we have identified the items by the same numbers as in Table 32 (which is not, of course, the order of their appearance on the questionnaires). It should be remembered that components analysis is designed to explain variance, and we should not expect it to highlight items that express the most universal attitudes of mountain elites. Indeed we are doing this in order to identify distinctive clusters of attitudes. The weighting of the factors yielded by this procedure tends to bring out the reverse of some of the points emphasized in Chapter 4 for precisely this reason—after the first factor the analysis increasingly picks out attitude clusters that constitute distinctive minority views.

In addition to identifying distinctive attitude clusters (factors), components analysis can also provide the basis for further study of relationships between where individuals stand on a particular attitude cluster and their other attributes—whether by direct use of factor scores or by use of an index suggested by loadings in the components analysis. We report in Appendix C on some multiple regressions in which we use factor scores as dependent variables. However, use of factor scores has a number of disadvantages. In the first place, they tell us nothing about position on an absolute scale but only state a relative position. For example, an individual may have a high factor score (relative position) on Factor 3 (critical view of local schools), yet he may be more defensive than critical; he is just less defensive than the average man. If a cluster of attitudes can be conceptualized in quasi-absolute terms, it may be quite as important to locate individuals or groups in those terms as to assess the significance of differences between them. Second, inevitably one interprets factors in

Download

Copyright Disclaimer:
This site does not store any files on its server. We only index and link to content provided by other sites. Please contact the content providers to delete copyright contents if any and email us, we'll remove relevant links or contents immediately.