Correlations
Clicking the correlation button
in the Project Data window will display the correlation menu. This is used to calculate the similarity of a selected row of data with all other rows in the selected list.
There are two available correlation methods:
- Pearson product-moment correlation coefficient: Finds the linear relationship of two rows by examining their tendency to increase and decrease together.
- Spearman's rank correlation coefficient: Finds the relationship of two rows by ranking the values in those rows, then examining the difference between the ranks. This method takes a bit longer than the Pearson correlation, but it is less affected by outliers.
Correlation results are stored in a temporary column in the Project Data table. To make the results a permanent part of the project, choose the Keep Previous Correlation. Also, a prompt to keep the previous correlation will appear when a second correlation is run. Note that when the project is saved, the previous correlation results will be saved as well, whether or not it was "kept". However, it will still be in a "temporary" state the next time the project is opened.
To delete "kept" correlations, simply select the correlation from the Remove Correlation submenu.