Dual scaling and correspondence analysis of preferences, paired comparisons and ratings

The methods of dual scaling (DS) and correspondence analysis (CA) are used extensively in marketing research for analyzing categorical data. When applied to preferences, paired comparisons and ratings, dual scaling has been contrasted with correspondence analysis, as if the two techniques operated differently on these data. In this note, we show that correspondence analysis provides exactly the same solution as dual scaling once the data have been transformed by a so-called “doubling” with respect to the respondents.