Unlike most other statistical techniques, factor analysis--which operates on a matrix of intercorrelated variables--does not distinguish between dependent and independent variables.
Accordingly, factor analysis is often an exploratory method that looks for patterns of relationships in advance of theory. However, it can also be used as a confirmatory method to test for relationships postulated in advance.
The most common form of factor analysis is "principal components" analysis, which analyzes all variation in the marix of variables, as opposed to "principal factors" analysis, which analyzes only that portion of the variation that the matrix of items has in common. (For a more detailed discussion go to the StatSoft Electoronic Textbook.)
In essence, principal components analysis identifies what is common to a set of variables. Many researchers use factor analysis interchangably with principal component analysis, and that is done here..
Factor analysis typically begins by assessing commonality within a set of variables and then proceeds to determine how subsets of variables within that set differ from one another.