Correspondence Analysis is a Multivariate technique for statistics. It was initially proposed by Hirschfeld. In the later stages it got developed by Jean Paul Benzercri. When we talk about its conceptual nature, it is pretty much similar to Principal Component Analysis. The only main difference being that it applies to categorical rather than continuous data, What is largely similar to Principal Component Analysis is that it offers a means and method for displaying or summarizing a set of data in the form of two dimensions graphical set.
There are certain prerequisites for Correspondence Analysis. The first one being that the data should be non-negative and on the same scale. The method has the peculiarity of treating the rows and columns equally. One must remember that Correspondence Analysis is a descriptive technique and it has the scope of getting applied o tables whether the descriptive statistics gets applied or no.
Since it is a descriptive analysis technique it has been designed to analyse simple two way and multi way tables which contain some measures of correspondence between the different rows and columns. The output that we get from this is very similar to the nature of the result that we would get by administering Factor Analysis. The correspondence analysis also permits you to find out the structure of the other categorical variables that are included in the table. One of the most easily seen this kind of table is the two way cross tabulation table.
In the most usual kind of correspondence analysis, we find that that the cross tabulation table of the frequencies first gets standardized. This is done with a purpose which is to ensure that the relative frequencies across the entire set add up to the whole number 1.0.
There are several different variations of Correspondence Analysis that are available for researchers. One of the common ones is Detrended Correspondence Analysis ( DCA ) and the Cononical Correspondence Analysis
(CCA). When the Correspondence Analysis gets extended into the categorical variables we call it the multiple correspondences Analysis.
When we talk of the specific application of CA in social sciences and its extension in the form of Multiple Correspondence Analysis it was initially confined to the country of France and was brought out to the rest of the world by Pierre Bourdieu when he applied it in his Sociology research.