Academic Statistics · Module 07 of 11

Correlation analysis

This module covers Pearson's product-moment correlation, Spearman's rho and Kendall's tau, partial correlation and the interpretation of correlation coefficients with confidence intervals. Correlation underpins regression, reliability and factor analysis, and confusing association with causation is a common reason for examiner corrections.

Watch the first lesson of every module free. The remaining lessons, exercises and live sessions open when you enrol.

What you will be able to do

  • Choose Pearson, Spearman or Kendall correlation from the data and the shape of the relationship
  • Check linearity and outliers with scatterplots before computing r
  • Interpret the direction, strength and confidence interval of a correlation and r-squared as shared variance
  • Use partial correlation to control for a third variable

Key terms

  • Pearson's r
  • Spearman's rho
  • Kendall's tau
  • Partial correlation
  • Coefficient of determination
  • Monotonic relationship
  • Spurious correlation

Hands-on practice

Produce a correlation matrix with scatterplots for a provided dataset in SPSS or R and report two correlations with 95% confidence intervals.

Common pitfalls this module fixes

  • Inferring causation from correlation; use causal language only where the design supports it.
  • Computing Pearson's r on a curved relationship; inspect the scatterplot and use a suitable model or Spearman's rho for monotonic data.
  • Interpreting statistical significance as strength; a small r can be significant in a large sample, so report the coefficient and its interval.