R-squared is biased downward in a regression suffering from serial correlation.

Question

Quizplus · Accepted Answer

R-squared is not directly affected by serial correlation; it measures the proportion of the variance for a dependent variable that's explained by an independent variable(s) in a regression model, but it does not account for the independence of the observations. Serial correlation primarily affects the standard errors of the regression coefficients, leading to unreliable statistical tests, not directly biasing the R-squared value.

R-Squared Is Biased Downward in a Regression Suffering from Serial