Correlation
Correlation matrix
Computes every pairwise association among several variables at once.
Variables
- VariablesContinuous · Ordinal — at least 2
Options
- Correlation typePearson’s r · Spearman’s ρ · Kendall’s τb
- AdjustmentHolm · Bonferroni · None
What it reports
Pairwise correlations across several variables; n and a CI for every cell.
scipy · statsmodels · Open the accuracy report
When does this apply?
For computing every pairwise association among two or more quantitative variables at once. There is no omnibus test — each cell carries its own coefficient, its own n and its own confidence interval, so none of them is promoted to a headline result. The p-values are adjusted for multiplicity, because ten variables means 45 simultaneous tests. The “Missing values” option genuinely changes the answer here: under complete cases every cell rests on one shared set of observations, while under pairwise deletion each cell uses the complete observations of its own pair.