Which statements are incorrect? The homoskedasticity assumption in time series regression suggests that the variance of the error term cannot be a function of time. In a regression model, if variance of the dependent variable, y, conditional on an explanatory variable, x, or Var(y|x), is not constant, the t statistics and the confidence intervals are both invalid no matter how large the sample size is. A variable is standardized in the sample by subtracting off its mean and multiplying by its standard deviation. In time series regressions, it is advisable to check for heteroskedasticity first, before checking for serial correlation. For a given significance level, if the calculated value of the Durbin Watson statistic lies between the lower critical value and the upper critical value, the hypothesis of no serial correlation is accepted. The variance of the slope estimator decreases as the error variance decreases. Whenever there is strong heteroskedasticity, it is preferable to use ordinary least square rather than Weighted least square, which may use a possibly misspecified variance function. A regression model contains homoskedasticity if the Breusch-Pagan test results in a large p-value. Covariance stationarity focuses only on the first two moments of a stochastic process