Home / Statistical Tools / Analysis Tools / Correlation and Covariance / Options
Options¶
Worked examples
The How-To pages build this chart step by step from sample data, one walkthrough per data layout the tool accepts. Start there if you would rather see a finished result than read the settings.
The three analyses share one dialog ("Correlation"). Which statistic you get is decided by the ribbon entry you launch, not by a dialog setting:
| Ribbon entry | Statistic | Worksheet |
|---|---|---|
| Pearson's Correlation (Normal) | Pearson correlation matrix with p-values | Correlation |
| Spearman's Correlation (Non-normal) | Spearman rank correlation matrix with p-values | SpearmanCorrelation |
| Covariance | Covariance matrix (no p-values) | Covariance |
Data Tab¶
- Data Types: a per-column dropdown to reclassify any column if it was detected wrong
- Data Columns (select at least 2): the columns to correlate. Numeric columns only (Continuous, Integer, Count); all eligible columns start selected. Selecting fewer than two shows "Please select at least 2 data columns for correlation analysis."
There is no GroupBy, frequency column, or overlay option for these analyses, and no task pane afterward.
How the data is aligned¶
Blank and non-numeric cells are removed from each selected column first, and then every column is cut to the length of the shortest one. A blank in the middle of a column therefore shifts that column's values against the others, so the pairs no longer come from the same rows. At least two values per column are required.
Reading the output¶
- With exactly two columns, a compact table shows the coefficient and its p-value. With more columns, a full matrix shows a p-value row and a coefficient row per variable, with one cell per pairing.
- p-values at or below 0.05 are shown in bold red. Values that cannot be computed show "NA".
- The covariance matrix includes the diagonal, which is each variable's own variance (the sheet notes this).
The formulas, including the t-test behind the p-values, are on Math Details.