Crossref Citations
This article has been cited by the following publications. This list is generated based on data provided by Crossref.
Schoen, Harald
and
Weins, Cornelia
2014.
Handbuch Wahlforschung.
p.
241.
Article contents
Party Preference and Attitudes on Political Issues: 1948–1951
Published online by Cambridge University Press: 02 September 2013
Extract
The widespread interest in understanding the political behavior of the American electorate has been served by a number of different methodological approaches. One of the most widely known and used is the analysis of aggregate voting statistics along the lines indicated by the work of Louis Bean. Major problems susceptible to study through aggregate behavioral measures include those related to the consequences of urbanization, population migration, drastic economic crisis or long-term change, or important international developments. Trends in voting behavior and regional differences in voting fall within this type of analysis, with census tract information and political sub-division voting statistics providing much of the relevant data for investigation.
A different approach to the study of political behavior is provided by the analysis of data on individuals, information pertaining to the behavior of identifiable persons. This approach allows flexibility in the ultimate units of analysis. The use of data on individuals not only provides such gross information as the percentages of people forming large groups (Democrats and Republicans, well-informed and poorly-informed) in the population, but it also provides information about various characteristics of these groups within the population.
- Type
- Research Article
- Information
- Copyright
- Copyright © American Political Science Association 1953
References
1 Lazarsfeld, P., Berelson, B., and Gaudet, H., The People's Choice (New York, 1949), p. 27Google Scholar.
2 Janowitz, Morris and Miller, Warren E., “The Index of Political Predisposition in the 1948 Election,” Journal of Politics, Vol. 14, pp. 710–27 (11, 1952)CrossRefGoogle Scholar.
3 Ibid.
4 The data used here are from the series of studies conducted by the Public Affairs Program of the University of Michigan Survey Research Center. For a somewhat different treatment of some of the same data, see Belknap, George and Campbell, Angus, “Political Party Identification and Attitudes Toward Foreign Policy,” Public Opinion Quarterly, Vol. 15, pp. 601–23 (Winter, 1951–1952)CrossRefGoogle Scholar.
5 Campbell, A. and Kahn, R. L., The People Elect a President (Ann Arbor, 1952)Google Scholar.
6 Survey Research Center, “Public Reactions to American Foreign Policy” (07, 1951)Google Scholar.
7 For the second study which we shall discuss (see below), the time of the “past” and “current” party preferences is changed to “usual voting behavior” and “vote in 1948,” and the same two-fold description of preference is used.
8 Three means of determining “relative support for the national administration” were tried: 1) the absolute difference between the percentage of the group membership which indicated support for the policy and the percentage of the group membership which indicated opposition, 2) an average rank ordering of the groups derived from combining the rank order of the groups in terms of the percentage supporting an issue within each group with the inverse rank of the groups in terms of the percentage of each group opposing an issue, and 3) the ratio of the number of group supporters to the total number of the group who took a stand on the issue (which could also be expressed as the ratio of the percentage who supported the issue to the total percentage who took a stand on the issue). Inasmuch as we were concerned only with those members in each defined group of party preference who actually expressed attitudes on the relevant policy items, method 3 was used throughout.
9 The possibility that the particular combination of the six available issues was accidentally responsible for the neatness of the array was examined. Various combinations of three or more issues were tested and all of them produced about the same distribution of groups. Some variations were observed in the central portion of the distribution, but it could be assumed that such variance was a product of the decreased reliability which resulted from decreasing the number of issues included in the index. The consistency with which any group of three or more issues produced the same array of political groups can, in fact, be exploited to demonstrate the usefulness of this approach to the analysis of both intra- and inter-political party shifts.
10 Cited above, n. 4.
11 Belknap and Campbell give major attention to the second of these hypotheses. Their discussion proceeds from the same general body of data which is under consideration here, i.e., data from the Survey Research Center, June, 1951 survey of the adult population.
12 See The People's Choice, p. xviii.
13 Although ND and NR scores on support for the administration in 1951 did tend to be more extreme than the scores of the DD's and RR's, the differences were not of a magnitude sufficient to preclude the possibility that sampling error or some other fortuitous factors produced them.
14 See notes accompanying Tables I and II on small cell populations.
15 The analysis of various combinations of three issues (instead of six), as described in footnote 9, was also made on the data from the 1948 study. The results were comparable in that any three issues whether they were three domestic, three foreign policy, or a combination of two of one and one of the other, produced similar orderings of party preferrer groups. Again, variations of rank ordering were observed in the middle-range groups—but it could be assumed that such variance was a product of decreased reliability which resulted from decreasing the number of issues included in the index.
16 For example, the DR. group from the first study and the DR. group from the second study were considered as one group of people. They had in common the experience of shifting from Democratic to Republican preference at some time prior to (or, conceivably, during) an SRC interview. Each member of the group had also responded to six items eliciting attitudes toward national governmental policy. The group had 6N opportunities (where N = group membership) to verbalize these attitudes (and thereby demonstrate political partisanship). The group score was computed as the number of pro divided by the number of pro and con responses. Again the DK and NA responses were not included in the scoring because of our interest in relating attitudes and party preference rather than in relating party preference to the absence of attitudinal commitments.
17 See n. 12.
- 1
- Cited by