Why it matters
Regression can summarise relationships and support prediction. Interpreting its output requires attention to units, assumptions and how the data were generated.
A worked example
A fitted model predicts sales = 100 + 3 × advertising spend. An extra unit of spend is associated with three extra sales in this model.
Illustrative example · simplified assumptionsA common mistake
Reading a fitted coefficient as a guaranteed causal return.
Where the idea needs care
Units, sample, functional form and omitted factors matter. A coefficient is not automatically causal, and extrapolation can fail.
Apply the idea
Explain this concept using a different example from your spending, work, business or a policy debate. State what stays fixed and what could change the result.
Sources and further study
Examples and explanations by Alibomics. Numeric illustrations are not current market quotations.