factor scores for all respondents on each of the five factors back into the
original data, in effect to create five new variables. These new variables then
provided the data basic for the description of right-wing opinions. Briefly, Lipset
and Raab proposed to organize radical opinion by crossing two dimensions--
Economic Conservatism and Intolerance. The factor analysis, it will be recalled, had given us three potential components for an ordinal representation
of the latter, and one for the former. Consequently, we assigned respondents
to one of five positions (categories) on an index of general intolerance as a
function of their position on the anti-Negro, anti-Semite, and cultural intolerance factors as measured by factor scores. An index of Economic Conservatism was similarly derived using the single factor so named.
This typology isolated four distinct groups of people: Radical Rightists, those
who had high scores on both the intolerance and economic conservative dimensions; Rednecks, those highly intolerant but not economic conservatives; the
Old Guard, economic conservatives but otherwise unprejudiced; and the Consistent Liberals, those who had very low scores on both the dimensions. This
typology, which corresponded to the initial theoretical differentiation of issue
publics suggested by Lipset and Raab before our factor analysis became the
base for their analysis in Chapter 11.
See especially, R. J. Rummel "Understanding Analysis," Journal of Conflict Resolution, IX ( December, 1967), pp. 444-480; Hayward R. Alker, "Statistics and Politics," in
S. M. Lipset, ed., Politics and the Social
Sciences ( New York: Oxford University Press, 1969), pp. 244-313; and H. H. Harmon, Modern Factor Analysis ( Chicago: University of Chicago
Press, 1960). This is the most comprehensive and rigorous discussion of
factor analytic techniques.
H. M. Blalock, Causal Inferences in Nonexperimental Research (Chapel
Hill: University of North Carolina Press, 1964), p. 169.
As an amusing illustration of this point, Price cites
J. S. Armstrong, "Tom
Swift and His Electric Factor Analysis Machine," American Statistician, XII( 1967),pp.17-21.
Harmon, op. cit., Parts II and III.
"An eigenvalue is the root of the characteristic equation [R - λI] = O,
where R is the correlation matrix, λ is an eigenvalue, I is an identity
matrix, and the brackets mean that the determinant is being computed." Rummel, op. cit., p. 466 n.
Johann Galtung, Theory and Methods of Social Research ( New York: Columbia University Press, 1969), Part II, Chapter 3. We highly recommend this chapter as a supplement to this discussion.
The latter, it should be noted, had been prepared by investigators chiefly
concerned with studying anti-Semitism.
We do not, as we have said, believe factor analytic techniques capable,