A Flexible Projective Technique Applied to the Measurement of the Self Images of Voters

Abstract

The difficulties of developing projective measures that are valid, objective, reliable, simple and convenient to administer, and flexible are briefly discussed. A technique specifically constructed to overcome these problems is described in the context of an experiment designed to measure selected dimensions of the self images of voters. This technique lends itself to modification so that it may be applied to a wide range of problems where projective measures are required that can be readily evaluated in terms of their validity and reliability. In the present study validity is demonstrated using the external criterion of actual voting habits.

Comparison of the preferred and rejected self images provides a measure of reliability similar to that obtained by using equivalent forms of the same tests. Product moment correlations for all eight dimensions are significant beyond the 5% level on a two-tail test.

Differences between the two groups of middle class voters are discussed in terms of the stronger “social conscience” of middle class Labour supporters. A hypothesis regarding the ambivalence of working class Conservatives is also considered.