Create a Correlation Matrix

The correlation matrix is a standardised variance-covariance matrix that emphasises the relative variation between two elements. Raw data is standardised by calculating the correlation coefficient so all elements have a variance of one and a mean of zero. The correlation coefficient is unit-less and is the ratio of the covariance of two elements to the product of their standard deviations. A correlation matrix can be calculated for either an entire dataset table or a subset.

Activity Steps

1. On the Discover ribbon tab, in the Tools group, select Correlation Matrix.

   The Correlation Matrix panel displays.

2. Select the Input Table.
3. Select the required element Fields:
4. Name the Output Table.
5. Select the Correlation Method from:
   • Pearson product-moment—Default
   • Kendall's Tau rank
   • Spearman's rank
6. Select whether to Calculate significance test for correlations.
7. Click Run.