¹ This paper is a reduced version of a considerably longer one (83 pages) entitled “An Examination of Voting Patterns in the 23rd and 24th Sessions of the General Assembly,” Research Report No.54, Dimensionality of Nations Project, University of Hawaii, 1971.Google Scholar The latter report includes: (1) complete information on the factor analyses, i.e., presentation of entire loadings matrices, etc., (2) separate analyses of the 23rd and 24th sessions by correlating the attribute dimensions with each of the votes, and (3) greater technical detail and justification. Copies of the paper may be obtained directly from author upon request.

I would like to thank David Davis, John Krupa, Kathleen Dul, and Mary Jo Snyder for their work as coders on the project; Myrtle Cassel (FAU) and Ora Barber (DON project) as typists; Carol Jones and Pauline Kartrude as programmers; and the FAU Research Committee for its support out of NSF monies.

² Vincent, Jack E., “Predicting Voting Patterns in the General Assembly,” American Political Science Review, 65 (1971) 471–498.Google Scholar

³ See Alker, Hayward R. Jr., and Russett, Bruce M., World Politics in the General Assembly (New Haven: Yale University Press, 1965);Google Scholar and Russett, Bruce M., “Discovering Voting Groups in the United Nations,” American Political Science Review, 60 (1966) 327–339.CrossRefGoogle Scholar

⁴ “Economic development,” etc., refer to “factor names” with specific meaning based on correlation with other variables. This point will be discussed further shortly.

⁵ This study, technically, is not a “replication” of the previous study in the sense that the same steps are gone through to see if the same result obtains. Rather, it is an effort to see, in spite of certain differences in procedures, if the “important predictors” of the previous study are also important for these data. The most important differences can be summarized as: (1) use of a second stage factor analysis in the first study, because of partially redundant factor scores (this point is fully explained in the previous study), (2) use of regression rather than mean estimates for missing data, and (3) use of General Assembly roll call and recorded votes of resolutions passed instead of all General Assembly and Committee roll call votes (successful Committee votes tend to closely parallel General Assembly votes). The above changes are viewed as “improvements” over the previous procedures, although, as will be seen, a similar result, in fact, obtains.

⁶ For a full treatment, see the various Dimensionality of Nations Research Reports.

⁷ Rummel, R. J., “Field and Attribute Theories of National Behavior: Some Mathematical Interrelationships,” Research Report No. 31, Dimensionality of Nations Project, University of Hawaii,” (1969), p. 6.Google Scholar The above and following brief introduction to social field and attribute theories also appears in Vincent, “Predicting Voting Patterns …” op. cit. Subsequent application should make clear some of the procedural implications of the above statements and the following formulae. Procedurally, the “test” of attribute theory is somewhat simpler than social field theory because the latter involves “distance” computations on the attribute dimensions and the determination of dyadic relations on the behavioral variables. In spirit, however, the two theories are extremely close in the sense that the primary guiding notion of both concerns “attributes in predicting behavior.” However, relations between dyadic partners is stressed in social field theory while relative location (to all states) on attributes is stressed in attribute theory.

⁸ Rummel, op. cit., p. 16.

⁹ Ibid., p. 7.

¹⁰ Ibid., p. 22.

¹¹ Rummel, , “Field Theory and Indicators of International Behavior,” Research Report No. 29, Dimensionality of Nations Project, University of Hawaii (1969), p. 10.Google Scholar

¹² When Rummel treats UN voting as behavior, he takes the Euclidean distances between the nations on the factor dimensions generated from an analysis of roll call votes. Ibid., p. 26C. Although this may allow treatment of UN voting data for certain purposes, nevertheless, as pointed out above, voting does not appear to be as dyadic in form in the same sense that exports, threats, etc., are.

¹³ Banks, Arthur S. and Textor, Robert B., A Cross Polity Survey (Cambridge: M.I.T. Press, 1963);Google ScholarRussett, Bruce M. et al. , World Handbook of Political and Social Indicators (New Haven: Yale University Press, 1964).Google Scholar See Vincent, “Predicting Voting Patterns …” op. cit., for exact sources of all attribute variables. Variable code numbers are retained in this study for convenience.

It is assumed that the factor score location of subjects will not change very much as resources are updated (not yet available). This is reasonable, particularly on the large factors, which have primary predictive importance in this study, because so many variables load heavily on such dimensions. Considerable changes would have to occur on a number of such heavily loading variables to get significant changes in factor score location.

¹⁴ Alker and Russett, op. cit., p. 30–31. These procedures do not typically put abstainers in the middle of the voting distributions. As explained by Alker and Russett, “Those abstaining against the pressure of a sizable majority come out closer to the scores of those who said No than they do to those in the affirmative.” Alker and Russett, op. cit., p. 31.

¹⁵ This places such states in the middle of the voting distributions.

¹⁶ These were Albania, Botswana, Gambia and Malta. Ukraine, Byelorussia, Madagascar, Mauritius, South Yemen and Swaziland were also dropped because of inadequate attribute information.

¹⁷ The judges were 44 upper division students at Florida Atlantic University who had just completed a course focusing on the United Nations. If the judges' most popular categories (positive, negative or neutral) are viewed as providing the “correct answers” and the judges' responses are then evaluated, Kuder-Richardson formula 20 shows a reliability of 81 for the “test.”

¹⁸ Baggaley, Andrew R., Intermediate Correlation Methods (New York: John Wiley and Sons, 1964), pp. 21–23.Google Scholar

¹⁹ See Harman, Harry H, Modern Factor Analysis (Chicago: University of Chicago Press, 1960);Google ScholarKaiser, Henry R., “The Varimax Criterion for Analytical Rotation in Factor Analysis,“ Psychometrika, 23 (1958) 187–200;CrossRefGoogle ScholarKaiser, Henry F., “Computer Program for Varimax Rotation in Factor Analysis,“ Educational and Psychological Measurements, 19 (1959) 413–430;CrossRefGoogle ScholarClyde, Dean J., Cramer, Elliott M., Sharin, Richard J., Multivariate Statistical Programs (Coral Gables, Florida: University of Miami, 1966) 15–19;Google ScholarVincent, Jack E., Factor Analysis in International Relations: Interpretation, Problem Areas and an Application (Gainesville: University of Florida Press, 1971);Google ScholarRummel, R. J., Applied Factor Analysis (Evanston: Northwestern University Press, 1970);Google Scholar and Vincent, Jack E., “Factor Analysis as a Research Tool in International Relations: Some Problem Areas, Some Suggestions and An Application,” Proceedings of the 65th Annual Meeting of the American Political Science Association (New York, 1969).Google Scholar

²⁰ Vincent, Factor Analysis in International Relations …, op. cit.

²¹ See: Hotelling, Harold, “Relations Between Two Sets of Variates,” Biometrika, 28 (1936), 321–377;CrossRefGoogle ScholarHotelling, Harold, “The Most Predictable Criterion,” Journal of Educational Psychology, 26 (1953), 139–142;CrossRefGoogle ScholarBartlett, M. S., “The Statistical Significance of Canonical Correlations,” Biometrika, 32 (1941), 29–38;CrossRefGoogle ScholarHorst, Paul, “Relations Among m Sets of Measures,” Psychomerika, 26 (1961), 129–149;CrossRefGoogle Scholar G. Thompson, “The Maximum Correlation of Two Weighted Batteries,” The British Journal of Psychology: Statistical Section, Part I (1947), 27–34;Google Scholar and Jack E. Vincent, Factor Analysis in International Relations …, op. cit.

²² The 23rd and 24th Sessions were separately factor analyzed and the resulting factor scores related to the attribute predictors. The basic results were not different from the “combined analysis” presented above, which is given because of its simplicity over a separate presentation of each session.

²³ See Vincent, Factor Analysis in International Relations …, op. cit.

²⁴ To simplify presentation, hereafter the term, “no” will be used to describe those who are “against the majority,” realizing that abstainers tend to be closer to “no” than “yes” when a large majority supports a successful resolution, which is typically the case.

²⁵ The square of each correlation gives the unique variance explained because both sets of variables are internally orthogonal. Therefore it is permissible to sum the squares to get the aggregate variance explained. See Vincent, Factor Analysis in International Relations …, op. cit. and Vincent, “An Examination of Voting Patterns …,” op. cit., on this point.

²⁶ That is, democracy predicts “negative” votes about as often as it predicts “positive” votes, when individual votes are analyzed. This is fully treated in the larger paper, “An Examination of Voting Patterns hellip;,” op. cit., obtainable from the author.

²⁷ As in the case of variables loading in a factor loadings matrix, each variable in a canonical analysis has some “weight.” To facilitate interpretation, however, smaller weights are usually ignored in discussion. In this connection, the square of the weight gives a good measure of its importance. For example, a weight of 70 (70² = 49) is four times as important as a weight half its size, i.e., 35² = .12. The square of a canonical loading, when the variables in the set considered (one side or the other) are orthogonal, is equal to the proportion of variance explained between the variable in question and “its” canonical variate scores.

²⁸ See: For Study 1, Vincent, Jack E., The Caucusing Groups of the United Nations: An Examination of Their Attitudes Toward the Organization (Stillwater: Oklahoma State University Press, 1965);Google Scholar Study 2, Vincent, Jack E., “National Attributes as Predictors of Delegate Attitudes at the United Nations,” American Political Science Review, 62 (1968) 916–931;CrossRefGoogle Scholar Study 3, Vincent, Jack E., “The Convergence of Voting and Attitude Patterns at the United Nations,” Journal of Politics, 31 (1969) 952–983;CrossRefGoogle Scholar Study 4, Vincent, Jack E., “An Analysis of Caucusing Group Activity at the United Nations,” Journal of Peace Research, 2 (1970) 133–150;CrossRefGoogle Scholar Study 5, Vincent, Jack E., “An Analysis of Attitude Patterns at the United Nations,” Quarterly Journal of the Florida Academy of Sciences, 32 (1969) 185–209;Google Scholar Study 6, Vincent, Jack E., Falardeau, Joseph, Schwerin, Edward, Bozeman, Barry, and Jednak, Robert,“Generating Some Empirically Based Indices for International Alliance and Regional Systems Operating in the Early 196's,” International Studies Quarterly, 15 (1971) 465–525;CrossRefGoogle Scholar Study 7, Jack E. Vincent, “Predicting Voting Patterns in the General Assembly,” op.cit; Study 8, Vincent, Jack E., “ Testing Some Hypotheses About Delegate Attitudes in the United Nations and Some Implications for the Theory Building,”Research Report No.52, Dimensionality of Nations Project, University of Hawaii (1971).Google Scholar

²⁹ The interpretation may be found in Studies 2, 4, 5 and 8.

³⁰ This statement is included because whenever I have used judges to assess the implications of votes and attitudes toward the United Nations I have found the above to be true. Whether it applies to all aspects of the organization (aspects I have not probed) or other international organization matters is the kind of investigation I would like to stimulate by the above formulation.

³¹ It is possible, of course, that no such assessment can be made.

³² These weights are to be viewed as a rough gauge of predictive importance regardless of the techniques employed. For example, in Study 6, the means of the 49 regional groups were compared with the known universe means for all states on the factor dimensions generated from a 74-variable factor analysis. “Economic development” generated significant differences for over three-fourths of the groups, “democracy” for over one-third, “U.S. relations” for slightly less than one-third. The rest of the factors, with one exception, “population rate increase,” were less important than “democracy” and have not showed the same degree of consistency across studies that is evidenced by the three predictors treated above. Only one predictor, however, “economic development” has shown complete consistency, in the sense of always emerging as important (in these studies).

³³ This and smiliar technical points are discussed at length in Vincent, Jack E., “Comments on Social Field Theory,” Research Report No. 58, Dimensionality of Nations Project, University of Hawaii (1972).Google Scholar

³⁴ A trace correlation can be viewed as measuring the overlap between two spaces. See Vincent, “An Examination of Voting Patterns …,“ op. cit.

³⁵ See the various DON research reports (University of Hawaii). Rummel has had more success, in this regard, with model II than model I, but, the testing of model II, because of the enormous data matrix that it generates (if all states are considered) is presently difficult to deal with. In this connection, Tong-whan Park believes that model I has more potential than has previously been realized. See Park, Tong-whan, “The Role of Distance in International Relations: A New Look at the Social Field Theory,” a paper presented to the International Studies Association (New Orleans, 1970).Google Scholar

³⁶ I have distinguished between national attributes and other variables, such as those relating to the psychological make-up of decision makers, which might be viewed as “personal attributes,” although the latter can be, with some exercise of the imagination, viewed as “national attribute” data. These kinds of “personal” measures, however, have not, to date, been included in field theory treatments.