Notes Flashcards
How do we interpret the model weights for a bivariate model?
A 1 unit increase in X is associated w/ a Beta hat change in the expected outcome, holding all else constant.
What is an interaction term?
Two independent variables multiplied together (for example tenure x collegrad)
For a multivariate model with interaction terms (for example: Yi = B0 + B1 tenure + B2 collegegrad + B3 tenure x collegrad) what do the coefficients for B1, B2 and B3 tell us?
B1 shows how tenure affects Y for non-college graduates.
B2 shows how college graduation affects Y when tenure is zero.
B3 indicates whether the effect of tenure on Y is different for college graduates compared to non-graduates.
- also B2 is the difference in intercept between collegegrad and non collegegrad and B3 is the slope.
Adjusted R2 formula
= ((1 - R2 unadjusted) * (n - 1)) / n-k-1
What is the F stat? (looks like F(5,68)
(q , n-k-1) where q is the # of restrictions, n is the number of observations, and k is the number of independent variables.
How do you interpret B5 female x Hispanic
The expected gender difference between men and women’s incarceration rates who are Hispanic vs those who are white, holding everything else constant.
What is the Model SS
% variation explained by the model
What is residual SS
% variation not explained by the model
What is Prob > F
This is essentially the p-value. In regression output, this value helps determine if the model as a whole is statistically significant. It shows the probability that the observed relationship between the predictors and the outcome variable occurred by random chance.
What does F stat model look like
A slide
How to calculate the F stat
F = ((R2 unadjusted - R2 adjusted)/k) / ((1-R2 adjusted) / n-k-1)
How do I find unrestricted R2
The R2 value from the unrestricted model
How do I find restricted R2
The R2 value from the restricted model
