Regression & Predictive Modeling

Slope

The slope tells you how much the predicted outcome changes for each one-unit increase in the input. A slope of 3 means that whenever x goes up by 1, the model predicts y goes up by 3; a negative slope means y is predicted to fall.

In a simple regression with one input, the slope is the regression coefficient — the single most useful number for understanding the relationship's direction and strength in real units (e.g. dollars of extra spending per extra year of age). It describes a predicted association, not a guarantee for any individual, and it only equals a causal effect under strong assumptions that observational data rarely meet.