FISHER Function
Returns the Fisher transformation of a correlation coefficient.
Category
Statistical
Available in
All versions, Microsoft 365
Returns
A number
Overview
FISHER converts a correlation r (between −1 and 1) into a value z = ½ × ln((1 + r) ÷ (1 − r)) that is roughly normally distributed. That makes it possible to build confidence intervals and tests for a correlation.
The usual recipe: transform r with FISHER, add and subtract 1.96 standard errors (1 ÷ √(n − 3)), then convert back with FISHERINV.
Syntax
Arguments
| Argument | Required? | What it does |
|---|---|---|
| x | Yes | The value to transform — usually a correlation; must be greater than −1 and less than 1. |
Examples
Example 1 — 95% confidence interval for a correlation
A survey of 28 customers found a correlation of 0.6 between satisfaction and spend.
| A | B | |
|---|---|---|
| 1 | Measure | Value |
| 2 | Correlation (r) | 0.6 |
| 3 | Pairs (n) | 28 |
| 4 | Fisher z | 0.6931 |
| 5 | Std error | 0.2 |
| 6 | Lower 95% | 0.2924 |
| 7 | Upper 95% | 0.7951 |
The true correlation is likely between 0.29 and 0.80. The interval isn't centred on 0.6 — correlations bunch up near ±1.
Example 2 — A single value
Fisher z for a correlation of 0.5.
Returns about 0.5493.
Common mistakes
- #NUM! x is −1 or less, or 1 or more.
- #VALUE! x isn't a number.
Notes

Author
Kyle Stott
Certified Microsoft Excel Expert
Kyle has worked professionally with Microsoft Excel for over a decade and has consulted on and taught Excel best practices to over 400 companies across 30+ countries.