satisfaction

#1
in order to run a simple linear regression and then analyze the results statistically as we did using the least squares approach, certain conditions must be satisfied. which of the following statements is NOT one of the conditions that must be satisfied:

a. distribution of Y_i values must be normally distributed.
b. distribution of epsilon_i values must be normally distributed.
c. distribution of X_i values must be normally distributed.
d. the variance sigma squared must be constant.
e. the epsilon_i error must be independent or at least uncorrelated.

ok, so i've narrowed it down to b and e, but i can't remember/figure out what epsilon is...any help would be much appreciated.
 

Dragan

Super Moderator
#2
in order to run a simple linear regression and then analyze the results statistically as we did using the least squares approach, certain conditions must be satisfied. which of the following statements is NOT one of the conditions that must be satisfied:

a. distribution of Y_i values must be normally distributed.
b. distribution of epsilon_i values must be normally distributed.
c. distribution of X_i values must be normally distributed.
d. the variance sigma squared must be constant.
e. the epsilon_i error must be independent or at least uncorrelated.

ok, so i've narrowed it down to b and e, but i can't remember/figure out what epsilon is...any help would be much appreciated.

The answer to your question is "c".

Look, in classical regression, X, the independent variable, is assumed to be Fixed - not random.


Note: epsilon is the (random) error term associated with the regression model.