Suppose you run a regression of an indicator for attendance in this course (1=at
ID: 3070391 • Letter: S
Question
Suppose you run a regression of an indicator for attendance in this course (1=attended, 0=did not attend) on an indicator for living within walking distance of the school (1=within walking distance, 0=not within walking distance) and the estimated coefficient on the walking distance variable was 0.11 with a standard error of 0.07. The null hypothesis that attendance is the same across groups:
a. Cannot be tested without a larger sample size.
b. Cannot be tested without knowing whether the errors are homoscedastic.
c. Cannot be rejected at the 95% significance level.
d. Can be rejected at the 95% significance level.
Explanation / Answer
Option b is the right one.
We cannot be tested without knowing whether the errors are homoscadastic or not. Because it is one of the assumptions regarding the regression analysis.