Determining correlation between two values
When you observe the data values in a table, you might want to determine whether there is significant correlation between numerical values in two columns. For example, when you view information about flights originating from a particular airport, you might be interested to know the correlation between departure delays and arrival delays.
Understanding the colors for correlation coefficients
When the correlation coefficients are displayed, the darkness of the background color indicates the strength of the correlation: The darker the color, the stronger the correlation.
A blue background color represents a positive correlation; a gray color represents a negative correlation.
Finding the correlation between selected columns or all columns
Follow these steps to display the correlation between selected columns:
If you are interested in the correlation coefficients for particular columns, follow these steps:
- Select the numeric columns for which you want to find the correlation coefficients.
- Right click the header on your selected column to display the column drop-down menu.
If you want to see the correlation coefficients for all columns, right click the table title bar to display the table drop-down menu.
Select Correlations… in the menu. The Correlation and Quick Aggregates modal window is displayed, which shows the correlation coefficient for each pair of selected columns.EXAMPLE: if you select the AirTime, Distance, and DepartureDelay columns (which must be numeric), you can determine the correlation between AirTime and Distance, between AirTime and DepartureDelay, and between Distance and DepartureDelay.
Finding the correlation between one column and all other columns
You can also determine the correlation between one particular column and all other columns. For example, if you are interested in how Distance is correlated with all other numeric columns in a table, follow these steps:
- Right click the column header to display the drop-down menu. For example, right click the Distance_integer column header.
From the drop-down menu, select Profile… to display a modal window similar to the following example:
In the Profile window, click Correlation to display the Correlation and Quick Aggregates modal window, as shown in the following example:
This example shows the correlation between Distance and other numeric values such as arrival delay, departure delay, and so on.