Why can an intercept confidence interval be scientifically unhelpful even when computed correctly?
x=0 may be far outside the observed predictor range
The intercept determines R-squared by itself
An intercept can never be estimated
Its standard error is always zero
Unlock full course notes by signing up
Inference in Linear Regression: difficulty 4
For n=20, a fitted slope is 0.75 and its SE is 0.25. Give (i) residual degrees of freedom and (ii) the t statistic for H0: beta1=0.