¹ The only systematic comparison of a national sample of these occasional activists to ordinary party identifiers that the author has been able to locate is Converse, Philip E. and Dupeux, George, “Politicization of the Electorate in France and the United States,” in Campbell, Angus, Converse, Philip E., Miller, Warren E., and Stokes, Donald E., Elections and the Political Order (New York: John Wiley and Sons, Inc., 1966), pp. 269–291 Google Scholar. Unfortunately, no attempt was made in this article to separate activists and ordinary members by party except through a correlation coefficient of the “party relatedness” of issues which obscured intraparty patterns of opinion.

² The earlier study, of delegates to the 1956 national conventions, was McClosky, Herbert, Hoffman, Paul J., and O'Hara, Rosemary, “Issue Conflict and Consensus Among Party Leaders and Followers,” American Political Science Review, 54 (06, 1960), pp. 406–427 CrossRefGoogle Scholar.

³ The actual wording of questions making up each scale is shown in Appendix A.

⁴ See Appendix A for a more complete description.

⁵ See Appendix A for a more complete description.

⁶ The test for significance of difference of the means used in this and other tables is the t-test—formula #13.5 in Blalock, Hubert, Social Statistics (New York: McGraw-Hill, 1960), p. 175 Google Scholar. Degrees of freedom computed from formula 13.6, p. 176. All tests are one-tailed. Since many differences in means regarded as meaningful by the author did not attain statistical significance because of small N's, a difference of .050 was as a rule of thumb taken to be the minimum size of a “meaningful” difference in the means. Whether or not differences formed any consistent pattern was also regarded as important.

⁷ The correlation statistic used in this article is Pearson's r. Although Pearson's r and partial correlations (also used in this article and based on Pearson's r) assume interval data, the data used in this article are predominantly nominal or ordinal. Use of Pearson's r and partial correlation is believed to be legitimate on three grounds. First, experience indicates that the use of measures generally regarded as more appropriate for ordinal data may produce results virtually identical to those derived from Pearson's r (see Nie, Norman H., Powell, G. Bingham Jr., and Prewitt, Kenneth, “Social Structure and Political Participation: Developmental Relationships, Part I,” American Political Science Review, 63 [06, 1969], pp. 361–378 CrossRefGoogle Scholar). Some of the more important correlations used in this article were also run against Kendall's tau and Spearman's r as a check, and no significant differences in results were found.

Second, the scope of interpretation of the data used in this article has been limited. It is obviously incorrect to make an exact statistical interpretation of Pearson's r computed on ordinal data—to say, for example, that a correlation of .10 between SES and welfare liberalism “explains” one percent of the variation in scores on the welfare liberalism scale.

Third, experience in the use of these measures indicates that it is proper to make the following kinds of statements about correlations and partial correlations computed on ordinal data:

(a) Comparison of the contribution of variables to variation in a dependent variable may be made. A statement that SES explains more of the variation in welfare liberalism than party would be legitimate (although one cannot say how much more).

(b) Statements about lack of relationship may be made. One can say that SES makes a meaningful contribution to welfare liberalism while party does not (although it must be recognized that the cut-off point of meaningfulness is somewhat arbitrary). The standard of meaningful correlation generally used in this article is .10.

⁸ The somewhat different picture presented by the 1956 election may have resulted from the Republicans' feeling that they had a sure thing that year, Eisenhower's limited attraction for the conservative wing of the party, and the outpouring of volunteer participation that Stevenson's candidacy seems to have produced.

⁹ Wilson, James Q., The Amateur Democrat (Chicago: University of Chicago Press, 1962), Chap. IGoogle Scholar.

¹⁰ McClosky et al., op. cit. There are a limited number of studies of various categories of party functionaries in geographically restricted areas. Most, unfortunately, do not include systematic comparison of the attributes of activists to the appropriate group of ordinary party members. One study that did make such a systematic comparison, in the Detroit area, was Eldersveld, Samuel J., Political Parties: A Behavioral Analysis (Chicago: Rand McNally, 1964)Google Scholar. Eldersveld found the same pattern reported by this article and by McClosky, although the same conclusions were not drawn. Other studies of opinions or social characteristics of party activists in limited geographic areas include Ippolito, Dennis S., “Political Perspectives of Suburban Party Leaders,” Social Science Quarterly, 49 (03, 1969) pp. 800–815 Google Scholar; Constantini, Edmund, “Intraparty Attitude Conflict: Democratic Party Leadership in California,” Western Political Quarterly, 16 (12, 1963), pp. 956–972 CrossRefGoogle Scholar; Salisbury, Robert H., “The Urban Party Organization Member,” Public Opinion Quarterly, 29 (Winter, 1955–1966), pp. 550–564 CrossRefGoogle Scholar; Ranney, Austin and Kendall, Wilmoore, Democracy and the American Party System (New York: Harcourt Brace and World, Inc., 1956)CrossRefGoogle Scholar.