Non-parametric correlation

The spearman correlation is an example of a nonparametric measure of strength of the direction of association that exists between two variables.

Other possible tests for nonparametric correlation are the Kendall’s or Goodman and Kruskal’s gamma.

Spearman’s correlation is therefore used to determine which relationship is monotonic.it examines if there exist a monotonic component of association between two ordinal or continuous variables.

Spearman’s correlation

Procedure

1. Click Analyze- Correlate-Bivariate
2. Drag the variables into the Variables box
3. Make sure to uncheck the Pearson checkbox( it’s always selected by default) and select the Spearman check box in the correlation coefficients area
4. Click the Ok button to view the results in the output window.

SPSS generates a single correlation table, which contain the spearman’s correlation, its significant value and the sample size. A high value of R_s, indicate a stronger the relationship between the variables.

NB/  check for monotonicity before running the spearman test, plot the data on a scatter plot.