Cronbach's α — Internal Consistency Reliability of a Scale
Assesses whether the items of a scale/questionnaire measure the same construct: enter the "subject × item" score matrix to compute Cronbach's α, plus the α with each item deleted (to identify items dragging down reliability). A routine metric in scale development and validation. Computed locally in your browser; data are not uploaded.
① Enter score matrix
One subject per row, with each item's score within the row (space- or comma-separated). All rows must have the same number of items. Reverse-scored items should be reversed first.
How to use & methodology
What is Cronbach's α and what counts as adequate?
It measures the internal consistency of a scale's items (whether they measure the same thing). Common standards: ≥0.9 excellent, 0.8–0.9 good, 0.7–0.8 acceptable, below 0.7 low. Higher (≥0.9) is usually required for individual diagnosis.
How is the data prepared?
Arrange it as a score matrix with one subject per row and one item per column, then paste it (space- or comma-separated). Reverse-scored items must be reversed first, or they lower or even produce a negative α.
How to use "α after deleting this item"?
If deleting an item raises the overall α, that item is poorly consistent with the others and may be revised or removed. But judge alongside content validity — do not delete items by the numbers alone.
What if α is low or negative?
Check for unhandled reverse items, whether the items measure the same construct, and whether the sample is too small. Multidimensional scales should compute α per dimension; a high α still does not mean unidimensional — run exploratory/confirmatory factor analysis when needed.