Correlation

If data is normally distributed (Parametric) conduct Pearson’s correlation,

For not normally distributed data (Non-Parametric) conduct Spearman’s rank correlation

• For both, the following steps are to be followed:
  • Analyze
  • Correlate
  • Bivariate/Partial/Distances
    • Bivariate for continuous variables
    • Partial for one continuous and another dichotomous variable
  • Move the variables to the variables tab
  • Select the type of correlations
  • Check the output table

Output:

• Sig. indicates whether the score would be correct 95% of the time for this sample or 99%.
  • If Sig. score is less than <0.05, it indicates,  the scores are significant at 95% confidence interval
  • If Sig. score is less than <0.01, it indicates  that the data is significant at 99% confidence interval
  • Once, significance is established, the Pearson/Spearman correlation score becomes important. That score indicates the strength of the relationship.
  • Scores closer to +/- 1 are strong and closer to 0 are weak. The scores can only be between -1 to +1