3
$\begingroup$

Consider a Simple Linear Regression with the following assumptions:

  1. The dependent variable is related to the independent variable and the error term like: $y = \beta_0 + \beta _1 x + u$

  2. We have a random sample of size $n$ following the population model in assumption #1

  3. The sample outcomes on $x$ are not all the same value

  4. $E[u|x] = 0$ is true

I am reading on the derivation of the sample variance for $\beta _1$. However, I don't understand why they assume that $\Sigma _ {i=1} ^ n x_i - \bar x$ is treated like a constant. Why is this true?

enter image description here

$\endgroup$
1
  • $\begingroup$ The condition |x is missing in the two variances in the middle of your equalities $\endgroup$
    – Bertrand
    Sep 25, 2022 at 13:17

1 Answer 1

1
$\begingroup$

The conditioning is on $x$, where $x$ represents all independent variables for all observations. Thus $x$ is treated as a constant throughout the derivation.

This is the standard method of deriving the variance of estimates, it is done conditional on the exogenous regressors.

$\endgroup$
1
  • $\begingroup$ That makes sense. Thank you! $\endgroup$
    – hu234
    Sep 26, 2022 at 20:20

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge that you have read and understand our privacy policy and code of conduct.

Not the answer you're looking for? Browse other questions tagged or ask your own question.