Factor analysis is a commonly used term in research and its main application is found in 1) reducing the number of variables. 2) Detecting the structure of the relationship that exists between the variables. It can also be called as classification of variables. It is often also termed as a data reduction or structure detection method. The Confirmatory Factor Analysis allows the testing of specific hypotheses for a particular set of variables in one or many samples. The correspondence analysis is a detective or exploratory technique that has been designed to analyse two way and multiway tables that contain some measures of correspondence that exist between the rows and the columns. The result gives the information which has similar characteristics to those that have been produced by factor analysis and they explore the structure of the categorical variables that are included in the table.
When Factor Analysis is used as a data reduction technique it helps to identify relevant data from a large data set. It can be understood by a simple example. Suppose the researcher wants to measure the height of 100 persons in both inches and centimetres. In this case the height would be measured in two variables. A future research, that is to identify the effect of different food supplements on height, both the measures of calculating height would not be required as height is only a single characteristic off a person, irrespective of how it is measured. So we would eliminate one variable by using factor analysis. Another example of factor analysis that is more relevant would be let’s take a study on the measure of the satisfaction of people with their lives. A satisfaction questionnaire is the fundamental thing that is designed where the focus is on asking questions about various things such as how satisfied they are with their hobbies, how much time to they devote to recreation etc. and the analysis of the responses would show a strong correlation with each other. And a high correlation would actually show that they are quite redundant and the responses can be clubbed into a single scatter plot and eventually the two variables can be reduced to a single factor for the study on satisfaction level.
Thus, factor analysis helps and aids data reduction and arrive at conclusions in a much easier way and arrive at more relevant conclusions.