Econometrics Exam 2
When plugging in an observation to the variables of a model, if it's log(y) = beta0 +
beta1*log(x1)... - ANS-Do loge^1(observation##) in your calculator.
Turning point of y-hat given x (based off beta coefficients) - ANS-If beta1 is positive and
beta2 is negative, the quadratic relationship has a downward opening parabolic shape.
the estimated change in y-hat for a given change in x - ANS-derivative of y-hat wrt x
multiplied by the change in x (so the effect of y of a change in x depends on the
beginning value of x as well as the magnitude of the change in x). SETTING THIS
EQUAL TO ZERO WILL GIVE THE TURNING POINT WHERE Y-HAT REACHES ITS
LOCAL MAX OR MIN
FORMULA: change in log(y) given change in x - ANS-100 * (e^(betacoeff) - 1) will give it
to you in % so you don't need to change it anymore I belive
Ceteris Paribus - ANS-partial effect of x on y can be found w derivative of y wrt x (and
all of the beta coefficients that x is a part of if there are interaction terms)
R^2 (unadjusted r-squared) - ANS-estimate of the proportion of variation in the
dependent variable explained by the independent variable (doesn't influence MLR4:
zero conditional mean, and isn't needed for an unbiased estimation)
adjusted r-squared (R^2) - ANS-A goodness of fit measure in multiple regression
analysis that penalizes additional explanatory variables by using a degrees of freedom
adjustment in estimating the error variance (but it is still a biased estimator of the
population R^2). = 1 - ( (1-r^2)(n-1) / (n-k-1) )
Robust standard errors - ANS-Standard errors of the estimated parameters of a
regression that correct for the presence of heteroskedasticity in the regression's error
term. So that we can get unbterm-7iased estimators
Without homoskedasticity, the estimators of the variances are... - ANS-biased, and the
t-stat and F-stat are not distributed within their distributions anymore
Breush-Pagan test for Heteroskedasticity - ANS-1. Estimate the model using OLS to
obtain the squared residuals (u-hat^2) by regressing y on all of the x vars
When plugging in an observation to the variables of a model, if it's log(y) = beta0 +
beta1*log(x1)... - ANS-Do loge^1(observation##) in your calculator.
Turning point of y-hat given x (based off beta coefficients) - ANS-If beta1 is positive and
beta2 is negative, the quadratic relationship has a downward opening parabolic shape.
the estimated change in y-hat for a given change in x - ANS-derivative of y-hat wrt x
multiplied by the change in x (so the effect of y of a change in x depends on the
beginning value of x as well as the magnitude of the change in x). SETTING THIS
EQUAL TO ZERO WILL GIVE THE TURNING POINT WHERE Y-HAT REACHES ITS
LOCAL MAX OR MIN
FORMULA: change in log(y) given change in x - ANS-100 * (e^(betacoeff) - 1) will give it
to you in % so you don't need to change it anymore I belive
Ceteris Paribus - ANS-partial effect of x on y can be found w derivative of y wrt x (and
all of the beta coefficients that x is a part of if there are interaction terms)
R^2 (unadjusted r-squared) - ANS-estimate of the proportion of variation in the
dependent variable explained by the independent variable (doesn't influence MLR4:
zero conditional mean, and isn't needed for an unbiased estimation)
adjusted r-squared (R^2) - ANS-A goodness of fit measure in multiple regression
analysis that penalizes additional explanatory variables by using a degrees of freedom
adjustment in estimating the error variance (but it is still a biased estimator of the
population R^2). = 1 - ( (1-r^2)(n-1) / (n-k-1) )
Robust standard errors - ANS-Standard errors of the estimated parameters of a
regression that correct for the presence of heteroskedasticity in the regression's error
term. So that we can get unbterm-7iased estimators
Without homoskedasticity, the estimators of the variances are... - ANS-biased, and the
t-stat and F-stat are not distributed within their distributions anymore
Breush-Pagan test for Heteroskedasticity - ANS-1. Estimate the model using OLS to
obtain the squared residuals (u-hat^2) by regressing y on all of the x vars
