Factor analysis has the trait of being able to simplify complex problems. In terms of the input and output of factor analysis, you enter much and yield little. In short, factor analysis brings in a large list of variables. After bringing in these variables, factor analysis finds the commonalities and groups the variables into factors. These factors can act as new variables, which include the original meanings of the initial variables.
Researchers often use factor analysis to create hypotheses about a set of data. In this case, researchers use factor analysis with an open mind; there are no assumptions about how each variable relates to each other. This use of factor analysis is often called "exploratory factor analysis," because it allows the researcher to explore relationships in the data. After the potential relationships are well-understood, the researcher can then formulate hypotheses about the phenomenon of interest.
While factor analysis has the characteristic of hypothesis creation, it also has the characteristic of theory checking; in other words, factor analysis is suitable in both the early stages and the late stages of an investigation. The theory-checking version of factor analysis often takes the name of "confirmatory factor analysis." This type of factor analysis allows a researcher to use the variables of a certain theory and the presupposed number of factors to perform factor analysis. Then, the researcher "confirms" the theory by making sure that the variables fit on these factors.
Many researchers employing factor analysis find that the process of factor analysis is similar to that of principal component analysis. However, there is one important characteristic belonging to factor analysis that principal component analysis lacks: interpretability. Although the factors of factor analysis are comparable to the principal components of principal component analysis, only the factors can be interpreted in normal language. This characteristic has helped factor analysis gain in popularity in scientific fields such as psychology in which descriptions of results help scientists apply the results.