Creating a correlation matrix for conjoint simulations
Creating a correlation matrix for your simulations is easy.
- Run your preference and revenue share simulations by inserting the product names, attributes and levels that you want.
- Click then .
- Now you can see a correlation matrix where the row and columns headers corresponds to the concepts you inserted earlier.
- You can optionally export your correlation matrix as
How to interpret correlation in conjoint simulations?
Conjoint.ly reports Pearson correlation coefficients for individual shares of different items. It means that correlations are based not on preferences for concepts, but on the shares of preference of concepts for each individual:
Working with shares of preference naturally leads to a bias towards negative values of the correlation coefficient. Therefore one must be careful not to put too much emphasis on low negative values of correlations.