Reason No. 1: We may prefer to report the usual OLS standard errors and test statistics unless there is evidence of heteroskedasticity.

• 理由1：除非有证据显示异方差存在，我们仍会偏好于常规OLS的标准差及检验统计量。