Which of the following lists the standard linear regression assumptions?

• A. Linearity, independence, homoscedasticity, normality of residuals, and correct model specification
• B. Linearity, independence, heteroscedasticity, normality of residuals
• C. Nonlinearity, dependence, constant variance of outcome, non-normal residuals
• D. Linearity, independence, homoscedasticity, multicollinearity

Answer: (A)

This question tests the set of assumptions that make ordinary least squares linear regression provide reliable estimates for inference. The standard list includes five elements: the relationship between the predictors and the mean of the outcome is linear; the observations are independent; the residuals have constant variance across all levels of the fitted values (homoscedasticity); the residuals are normally distributed (for valid inference in small samples); and the model is correctly specified (relevant variables and the right functional form are included).

Linearity ensures the model y = β0 + β1x1 + ... + ε captures the true relationship in a straight-line sense with respect to the parameters. Independence means each observation comes from a separate, unrelated unit, so errors aren’t correlated across observations. Homoscedasticity means the spread of the residuals does not change with the level of the fitted value, so standard errors are trustworthy. Normality of residuals supports the validity of t-tests and F-tests used to assess hypotheses about the coefficients, especially with smaller samples. Correct model specification requires including the right variables and using the appropriate functional form so the estimated effects are unbiased and meaningful.

options that mention heteroscedasticity, nonlinearity, dependence, or multicollinearity mix in conditions that are not part of the standard assumptions (heteroscedasticity is a violation, nonlinearity and dependence are not the assumed form, and multicollinearity is not an assumption but a potential issue affecting precision). The listed correct set captures the essential requirements for reliable OLS inference.