We have used a number of statistical measures of social, psychological, and political phenomena in this book. In this appendix we discuss the construction of these measures, in the order in which they appear in the text.
The PPS is a Guttman-type, cumulative scale of political participation constructed, for white respondents, from questions 1, 2a, 2c, 2e, 3, 4, 9 on political discussions; 27 on voting; 6, 7, 8 on campaign participation; and 5, 5a, 11, 11a on officeholding and political memberships. The comparable questions for Negro respondents are 1, 2a, 2c, 2f, 3, 4, 9; 30; 6, 7, 8; and 5, 5a, 11, 11a.
Respondents were dichotomized into those who had ever engaged in any of these activities and those who had not. The percentages of the Negro and white southwide samples with positive responses on each item were as follows:
Held office or belonged|
to political group
There is no general agreement about what constitutes a satisfactory cumulative scale. Guttman has proposed a Coefficient of Reproducibility (CR), which is merely the proportion of all responses falling into "pure" scale pat-