ABSTRACT
Factor analysis is based on the assumption that all variables are correlated to some degree. Therefore, those variables that share similar underlying dimensions should be highly correlated, and those variables that measure dissimilar dimensions should yield low correlations. Using the earlier example, if the researcher intercorrelates the scores obtained from the 30 personality tests, then those tests that measure the same underlying personality dimension should yield high correlation coefficients, whereas those tests that measure different personality dimensions should yield low correlation coefficients. These high/low correlation coefficients will become apparent in the correlation matrix because they form clusters indicating which variables “hang” together. For example, measures of ethnocentrism, authoritarianism, and aggression may be highly intercorrelated, indicating that they form an identifiable personality dimension. The primary function of factor analysis is to identify these clusters of high intercorrelations as independent factors.
