Bivariate correlation
Measure how strongly two or more numeric columns move together.
When to use it
Use a correlation when you want the strength and direction of a relationship, not a prediction equation. For example: do students who study more hours also have higher exam scores, and do more anxious students have lower scores?
Pearson’s r describes a straight-line relationship. Spearman’s ρ and Kendall’s τ are method options on this same page. They rank the scores first, so a curve or a few extreme scores moves them less.
Assumptions
Pearson’s r assumes a straight line and is pulled around by outliers. Spearman and Kendall do not assume a straight line. Tensr does not test normality or linearity inside this procedure.
Running it in Tensr
Options
Prop
Type
Reading the output
Hours, anxiety, and exam score for the 96 students. Flag significant is off, so the coefficients are printed without stars. Hours was built to rise with score. Anxiety was built to fall with it. Neither link is tight.
| hours | anxiety | score | |
|---|---|---|---|
| hours | 1 | -0.067 | 0.327 |
| anxiety | -0.067 | 1 | -0.374 |
| score | 0.327 | -0.374 | 1 |
| hours | anxiety | score | |
|---|---|---|---|
| hours | — | .516 | .001 |
| anxiety | .516 | — | < .001 |
| score | .001 | < .001 | — |
Pearson correlations among 3 numeric variable(s), 96 complete rows. Pearson correlations among 3 numeric variable(s), 96 complete rows. Metrics: Method = Pearson; Complete rows = 96; Significance α = 0.05.
Reporting (APA 7)
Pearson correlations among 3 numeric variable(s), 96 complete rows. Report the estimate in the table. This procedure is not summarised by one p-value.
Coming from SPSS
Analyze → Correlate → Bivariate.
SPSS puts Pearson, Spearman, and Kendall in one dialog, which matches this page. SPSS can also print a scatterplot from Chart Builder. Tensr’s chart switch is the shared include_charts option.
Related
One relationship with other columns held constant is partial correlation. Predicting a score from several columns is linear regression.