Statistics • Multiple Regression
s_e is the standard error of estimate and summarizes the typical size of residuals in the units of the response variable y. Smaller s_e generally indicates predictions are closer to observed values, holding the model form and predictors fixed.
The calculator uses the residual sum of squares (SSE) and the residual degrees of freedom to compute s_e = sqrt(SSE / df). The df depends on the sample size and the number of estimated parameters (including the intercept if selected).
Rows with missing or non-numeric values in the selected response or predictor columns are dropped so the regression and residuals are computed on valid numeric data. This prevents errors and keeps SSE and s_e consistent with the chosen variables.
It builds a sequence of models by adding predictors in the current checklist order and computes s_e at each step. This helps you see how the typical error changes as more predictors are included.
The histogram shows the distribution of residuals with a normal overlay, while the guides at plus/minus s_e and plus/minus 2 s_e provide practical error-size reference points. The strip plot groups observations into bands such as |e| < s_e, |e| < 2 s_e, and |e| >= 2 s_e to highlight small versus large errors.