How to interpret the correlation coefficient in empirical research
The correlation coefficient summarizes the direction and intensity of the relationship between two variables, but does not explain the cause itself.
Key information
- ✓The coefficient sign indicates the direction and the value absolute strength of the bond.
- ✓The correlation does not prove a causal relationship.
- ✓A back-up observation can greatly alter the outcome.
- ✓Statistical relevance does not automatically imply a high level of practical relevance.
What is the value of the coefficient
The coefficient takes values from negative to positive. The sign indicates whether the growth of one variable is usually accompanied by a decrease or a rise of the other. The assessment of dependence should be based on the convention applicable to the discipline and not on a random table from the Internet.
Pearson and Spearman
Pearson describes a linear dependence of quantitative variables under certain assumptions. Spearman uses ranking and assesses a monotonic dependence. However, it is not an automatic solution to every problem with distribution, so the choice must be justified by the characteristics of the data.
Figure and observations left out
Before reporting, prepare a spreadsheet chart, which may reveal a curve dependence, aggregate, or single result-driven observations.
How to Write an Interpretation
Give the type of correlation, the number, the value of the coefficient and the level of p, and then describe the direction and strength of the correlation.
Summary
The correlation should be read together with the chart, sample size, p-value and the context of the study.
Frequently asked questions
When should we choose Spearman over Pearson?
Spearman is useful, among other things, for order data and monotonic dependencies.
Need support with your work?
Describe the topic, scope and deadline. We will review the materials, recommend the right support and prepare a clear quote.
