Linear regression finds the best-fit line through a dataset: y = mx + b.
R² close to 1 = strong model fit. R² near 0 = poor fit.
The slope (m) indicates how much Y changes per unit increase in X.
Use this statistical calculator when you need a rigorous, reportable result — not a rough estimate. The formula is standard, but manual computation on large datasets introduces errors that undermine any subsequent analysis.
Choosing population standard deviation when the data is a sample is the most common statistical error with this type of calculation. Unless your dataset covers every possible member of the group, use sample SD (with n−1 denominator) for valid inference.
A quality engineer analyses the diameter measurements of 25 machined parts: mean 50.02 mm, standard deviation 0.08 mm. With a tolerance of ±0.25 mm, the process is well within specification — and the SD confirms the manufacturing variation is low enough that virtually no parts will fall outside tolerance.
Linear regression finds the line of best fit: y = mx + b, minimizing the sum of squared residuals.