Definitions
Another name for Y.
What is the dependent variable?
Interpret: Yi = B0+B1Xi+ui
What is the linear regression model with one regressor?
The two conditions for omitted variable bias.
Corr(X,Z) != 0
Corr(u,Z) !=0
This is when the error term is correlated with the regressor of interest.
What is endogeneity?
Interpret: B1 = 2 when X and Y are continuous variables.
What is a one-unit in X yields a two-unit increase in Y?
What are homoskedastic errors?
The extent to which a causal relationship is justified.
What is internal validity?
Interpret B1 = 3 in a multivariate model when X and Y are continuous variables.
What is a one-unit increase in X has an expected 3-unit increase in Y, holding all else equal?
This is how you would treat the null hypothesis at the 5% alpha level if your p-value is 0.02.
What is reject the null hypothesis?
This is when at least one regressor is a perfect linear combination of at least one of the other regressors.
What is perfect multicollinearity?
Interpret: R2=0.70
What is the regressors explain 70% of the variance in Y?
The four key assumptions of OLS.
1. The conditional distribution of ui given Xi is mean-zero
2. i.i.d.
3. Large outliers are unlikely
4. No perfect multicollinearity
This when a problem can only be solved numerically.
What is no closed form solution?
Interpret: B1 when Y is binary.
What is the change in the predicted probability that Y=1 associated with a unit change in X?
The equation for adjusted R2.
1-(n-1)/(n-k-1)*(1-R2)