Path: Janda's Home Page > Workshop > Factor Analysis > SPSS outline

In both tables, all variables with loadings (i.e., correlations) of .60 or more with a factor are in boldface. In the right-hand table, I've deleted all loading lower than .30 to clarify the structure.

The shaded matrix shows that the first six issues all correlate above .60 with the principal component of the original correlation matrix. However, four issues correlate above .60 with the next factor, which is unrelated to the first. These findings disconfirm unidimensionality in the parties' issue positions.

Rotation of the factor matrix, furthermore, shows that eleven of the thirteen issues breakdown into three disctinct clusters of variables, which I have named Economic Leftism, Cold Warism, and Liberalism. One variable relates about equally to all three factors, while another variable stands largely alone.

For more information about this analysis of party positions, see Kenneth Janda, Political Parties: A Cross-National Survey (New York: Free Press, 1980), Chapter 6 and especially pages 147-149.

Component Matrix (Unrotated factors)
Rotated Component Matrix

1
2
3
4
1
2
3
4

ownership of production

0.85
0.23
-0.24
-0.10
0.90

govt role in economic planning

0.81
0.22
-0.34
-0.13
0.89

redistribution of wealth

0.78
0.07
-0.19
-0.30
0.83

social welfare

0.74
-0.16
0.32
-0.10
0.82

secularization of society

0.73
0.30
-0.35
0.00

0.79

support of the military

0.63
-0.27
0.17
0.36

-0.75

alignment with east-west blocs

0.55
0.54
0.41
-0.12

0.74
-0.30

anti-colonialism

0.50
-0.78
0.16
0.05

-0.73
0.53

supranational integration

-0.05
0.76
0.14
0.17

0.83

national integration

-0.45
0.70
0.05
0.08
0.45

0.66

electoral participation

-0.24
-0.68
-0.12
-0.38

0.60
0.40

protection of civil rights

0.48
-0.08
0.65
0.18
0.44
0.48
0.52

interference with civil liberties

0.22
-0.21
-0.39
0.78

0.92

Meaning attached to rotated factors:

Economic Leftism
Cold Warism
Liberalism
?