Pearson correlation matrix, significance testing and Mantel-style correlation network. Upload or paste your data, select variables, run the analysis and customize a publication-ready figure.
Appearance Settings
Font size
Appearance settings affect the interface and result tables. Figure fonts and colors can also be customized separately below.
1. Add your data
Use a header row and one observation per row. Numeric columns are available for analysis. For best results, use at least 5 observations and avoid constant variables.
Supported: CSV, TSV, TXT and Excel (.xlsx/.xls). Excel reading loads SheetJS from cdnjs. Data are processed locally in your browser; they are not sent to a server.
2. Select variables and analysis options
Correlation matrix variables
Select numeric variables to display in the upper-triangle matrix.
Network variables
Select additional numeric variables to appear as nodes on the left, linked by slightly curved lines to the diagonal of the matrix. A variable can belong to either group.
Network method: each network edge represents a Mantel-style test between two Euclidean distance matrices (one built from each selected variable). The Mantel statistic is the Pearson correlation between the upper triangles; significance is estimated by permuting sample labels. This is not a multivariate Mantel test across a group of variables.
3. Analysis results
Correlation matrix
Network test results
4. Publication-ready figure
The figure is laid out on an A4 page (choose landscape or portrait below). An upper-triangle correlation matrix is combined with a Mantel network: each network variable is a node on the left, linked by curved lines to the diagonal cell of each matrix variable. Line colour shows the permutation p-value, line width shows the Mantel r class, and dashed grey lines are not significant.
Tip: drag a network variable (its red dot or name) on the figure to move it and change the distance between variables. Flip, curvature and spacing are in the panel below.
5. APA 7 interpretation of tables and figure
Each table and each portion of the figure is interpreted below. Statistics update with your data and settings. Downloads require login.
How to interpret the figure
Matrix squares: colour and square size represent the correlation coefficient (red = positive, blue = negative by default). Stars mark significance (* p < .05, ** p < .01, *** p < .001). Coefficient values can be switched on in the customization panel.
Network lines: connect each network variable (left) to the diagonal cell of a matrix variable. Line width encodes |Mantel r| (< 0.25, 0.25 – 0.5, ≥ 0.5).
Line colour: orange = p < 0.01, purple = 0.01 – 0.05, dashed grey = p ≥ 0.05 (not significant).
Printing: the figure is built at A4 proportions. Use the A4 PNG download (150, 300 or 600 dpi) or the print button, and choose margins “None” and scale 100%.
Mantel permutation p-values are empirical and depend on the number of permutations. With 499 permutations, the smallest attainable p-value is 1/500 = .002. Increase the permutation count for final reporting. Pairwise deletion can lead to different sample sizes across matrix correlations; the Mantel network uses complete paired observations for each variable pair.