Correlated Variables (SPSS)

In a prior article, I discussed correlation as it pertained to the “R” platform. In this entry, I will demonstrate how to generate a similar summary within the “SPSS” platform. We will be utilizing the data from the previous exercise for our example.

From the “Analyze” option on the top menu, select “Correlate” and then “Bivariate”.

Select the variable that will be analyzed. There is no need to select any additional options.

This produces the output below.

Typically, Pearson Correlation (“r”) values are interpreted as follows:

No Association: r value of 0

Small Association: r value of .1 to .3 (or) -.1 to .3

Medium Association: r value .3 to .5 (or) -.3 to -.5

Large Association: r value of .5 to 1 (or) -.5 to -1

Perfect Association: r value of 1

A rough rule of thumb is that only variables that possess a large association should be further examined within the context of the model.

That’s all for now, Data Heads! In the next article, we will discuss the ROC Curve and how it pertains to logistic regression.