Research Workbench

Cronbach alpha from item covariance

Estimate internal consistency from a complete item covariance matrix.

Your entries stay in this page’s memory and are not submitted to us. Downloads are generated locally. Reloading or leaving the page clears your entries. Do not enter patient information.

Square covariance matrix, items in the same row/column order

Use 2–20 items and a symmetric positive-definite sample covariance matrix. Reverse-key items before computing the covariance matrix.

The starting values are hypothetical. Results update as you edit. Use aggregate or synthetic inputs only.

Calculation results

Raw Cronbach alpha: 0.75
Total-score variance: 6
Items: 3
Interpretation note: Alpha alone does not establish a valid one-dimensional scale.

How to interpret the output

Alpha=k/(k−1)×(1−sum(item variances)/variance of total score). A high alpha does not establish unidimensionality or validity and can reflect redundant items. The reliability interpretation requires measurement assumptions; negative alpha flags an incompatible covariance pattern or coding issue, not a value to truncate.

Method reference

Source checked 2026-11-06. Read the reproducible calculator validation report.