Hostname: page-component-7479d7b7d-c9gpj Total loading time: 0 Render date: 2024-07-13T21:55:27.087Z Has data issue: false hasContentIssue false
  • English
  • Français

A Test of the Uni-dimensionality of Various Political Scales through Factor Analysis: A Research Note

Published online by Cambridge University Press:  10 November 2009

M. James Dunn ,
Rodney Schneck  and
Jan Lawson
M. James Dunn
Affiliation:
University of Alberta
Rodney Schneck
Affiliation:
University of Alberta
Jan Lawson
Affiliation:
University of Alberta
Get access Rights & Permissions [Opens in a new window]

Extract

Contemporary political science is increasingly interested in the problem of empirical measurement and at various times researchers need scales to measure political attitudes and behaviour. In this process the Canadian political scientist is faced with two alternatives: (1) he can develop new scales, or (2) he can use existing measures from the literature which in most cases mean using scales empirically developed and used in the United States.

Type
Notes
Information
Canadian Journal of Political Science/Revue canadienne de science politique , Volume 6 , Issue 4 , December 1973 , pp. 664 - 669
Copyright
Copyright © Canadian Political Science Association (l'Association canadienne de science politique) and/et la Société québécoise de science politique 1973

Access options

Get access to the full version of this content by using one of the access options below. (Log in options will check for institutional or personal access. Content may require purchase if you do not have access.)

References

1 In this study, uni-dimensionality is equated with homogeneity.

2 See Upshaw, Harry S., “Attitude Measurement,” in Blalock, Hubert M. and Blalock, Ann B., Methodology in Social Research (Toronto, 1968), 60–108Google Scholar; Robinson, John P., Rush, Jerrold G., and Head, Kendra B., Measures of Political Attitudes (Ann Arbor, 1968), 16–18Google Scholar; Scott, William A., “Attitude Measurement,” The Handbook of Social Psychology, 2nd ed.Lindzey, Gardner and Aronson, Elliot (London, 1968), 254–6.Google Scholar

3 The scales were “Canadianized” by the authors. For example, Washington was replaced by Ottawa or American was replaced by Canadian.

4 Variables 1, 2, 6, and 7 are from a questionnaire called “Political Attitudes Study,” from the Survey Research Center, University of Michigan. Variables 4 and 5 are from Trow, Martin, “Small Businessmen, Political Tolerance, and Support for McCarthy,” American Journal of Sociology, 64 (1958), 270–81CrossRefGoogle Scholar. Variables 3, 8, and 9 have been adapted from Robinson, Rusk, and Head, Measures of Political Attitudes, 70–202.

5 The items in this scale are from a questionnaire called “Political Attitude Study,” from the Survey Research Center, University of Michigan. Also see Stauffer, Samuel A., Communism, Conformity, and Civil Liberties (Garden City, New York, 1955).Google Scholar

6 This scale was developed by the authors.

7 Variables 19, 20, 21, 22, and 23 are from Horton, John E. and Thompson, Wayne E., “Powerlessness and Political Negativism: A Study of Defeated Local Referendum,” The American Journal of Sociology, 67 (1962), 489.CrossRefGoogle Scholar

8 Variables 24 and 25 were developed by the authors, variables 26, 27, and 28 were modifications from Woodward, Julian L. and Roper, Elmo, “Political Activity of American Citizens,” American Political Science Review, 44 (1950), 872–85CrossRefGoogle Scholar. Variable 29 is from Horton and Thompson, “Powerlessness and Political Negativism,” 489.

9 “Measures of Test Homogeneity,” Educational Psychological Measurement, 20 (1960), 751.

10 Robinson, Rusk, and Head, Measures of Political Attitudes, 16–18.

11 Once the sample had been determined and the questionnaires administered the data were subjected to a coding procedure. This was completed and the coded data were introduced into an SPSS computer program for factor analysis. The specific factor analysis program utilized was PA 2 which is principal factoring with iteration. Program PA 2 automatically replaces the main diagonal elements of the correlation matrix with communality estimates. Initial estimates of the communalities are given by the squared multiple correlation between a given variable and the rest of the variables in the matrix. PA 2 then employs an iteration procedure which continues until the difference between two successive communality estimates is negligible. See Nie, Norman H., Bent, Dale H., and Hull, C. Hadlai, Statistical Package for the Social Sciences (New York, 1970), 208–44.Google Scholar

12 Since items 4 and 5 do not load significantly on any factor they were deleted.