Deck 11 – Simple Linear Regression (LM10

Why does heteroskedasticity make OLS significance tests unreliable?

It biases standard errors, making $t$-statistics unreliable.

What is the key consequence of omitted variable bias in simple regression?

$hat{b}_1$ becomes biased and inconsistent.

How are total, explained, and unexplained variation related in regression ANOVA?

$\mathrm{SST}=\mathrm{SSR}+\mathrm{SSE}$.

What does $R^2$ tell you in simple linear regression?

The proportion of variance in $Y$ explained by $X$.