Statistical & Methods QA · DASS

APA 7 Regression Results Sentence Formatter

Create a two-line regression result that separates overall model fit from the selected predictor's coefficient.

Want to run the whole analysis in R, not just format the result? Core Statistical Analyses workflow ($39) ↗

How to report regression results in APA 7

A clear regression report separates the overall model from individual predictors. The model line gives R2, F with numerator and denominator degrees of freedom, and the model p value. The coefficient line gives the unstandardized coefficient b, its standard error, the standardized Greek β, t with degrees of freedom, and its p value.

Italicize Latin-letter symbols such as R, F, b, t, df, and p. Do not italicize SE or the Greek β. Because R2, standardized β, and p are bounded, omit their leading zeros. Retain the leading zero for F, unstandardized b, SE, and t, all of which can exceed 1. Exact probabilities use three decimals, with p < .001 replacing p = .000. Frequent mistakes include reporting β without b, omitting one of the model degrees of freedom, and confusing R2 with adjusted R2. If you report adjusted R2, label it explicitly in your surrounding prose.

Worked example

A three-predictor model explained 28% of the variance, F(3, 96) = 12.41, p < .001. Study time had b = 0.48, SE = 0.12, β = .39, and t(96) = 4.02.

R2 = .28, F(3, 96) = 12.41, p < .001.
Study time: b = 0.48, SE = 0.12, β = .39, t(96) = 4.02, p < .001.

Common questions

What goes on the model line and what goes on the coefficient line?

The model line gives R², F with numerator and denominator degrees of freedom, and the model p. The coefficient line gives the unstandardized b, its standard error, the standardized β, t with its degrees of freedom, and p.

Which symbols are italicized?

Italicize Latin-letter symbols such as R, F, b, t, df and p. Do not italicize SE or the Greek β.

Which values keep the leading zero?

Omit it for R², standardized β and p, which are bounded. Keep it for F, unstandardized b, SE and t, which can exceed 1.

How many decimals for p?

Exact probabilities use three decimals, with p < .001 replacing p = .000.

What mistakes should I avoid?

Reporting β without b, leaving out one of the model degrees of freedom, and confusing R² with adjusted R². If you report adjusted R², label it explicitly.