Econometrics Midterm || with 100% Error-free Answers.
What is econometrics? correct answers Tools of economic theory, math, and statistical inference
are applied to the analysis of economic phenomena.
How do we do econometrics? What are the steps involved? correct answers 1. Make a hypothesis
2. Collect data
3. Specify the model
4. Estimate the parameters
5. Check model specification - is the model good?
6. Test the hypothesis
7. Use the model to predict or forecast or answer questions
Simple linear model correct answers yi= B1 + B2Xi + ui
Bs are parameters/coefficients, X is regressor, u is error term
What is the error term? Why do we have an error term? correct answers omitted explanatory
variables,
aggregation of variables (consumption function)
model misspecification
may be determined by expectations rather than actual values
functional misspecification
x instead of x^2
measurement error
Linear correct answers linear in parameters:every term consists of a straightforward parameter
multiplied by a variable
linear in variables: every term consists of a straightforward variable multiplied by a parameter
when to reject the null correct answers reject: t-stat > t-critical
*don't forget to explain why: explanatory variable seen as significant at the % level
B2hat - t-crit*se(B2hat)< B2hat < B2hat+ t-crit*se(B2hat)
Method of OLS correct answers minimize the sum of the squared errors in order to have a very
fitted line
What is OLS? correct answers ordinary least squares
most popular application of regression analysis
What does OLS do? correct answers we sum the squared errors
Why do we use OLS? correct answers ◦ OLS is simple to do (compared to other regression
techniques)
◦ OLS has some nice properties - the regression line goes through the means
of X and Y and the sum of the residuals is zero (or very close to zero).
, ◦ OLS is the BEST estimator under a certain set of assumptions
How do we estimate the coefficients using OLS in a two variable model? correct answers
B2=sum(Xi-Xbar)(Yi-Ybar)/
sum(Xi-Xbar)^2
B1= Ybar - B2Xbar
- Be able to explain the intuition correct answers the goal is to minimize the sum of the squared
errors
Explanatory variable correct answers same as independent variable and regressor
Dependent variable correct answers Y or what is being explained by regressing data
Be able to draw a graph of a regression model and label all the parts (like the coefficients, the
error terms, etc) correct answers (look up in notes)
Causation and correlation correct answers Example:
A one dollar increase in income suggests a 51 cent increase in average
consumption
◦ Suggests or is correlated with (not causes)
Scaling and units and interpretation correct answers
Demeaning - what is it? Why do we do it? Interpret coefficient. correct answers We regress Y on
X* when X*=Xi-Xbar
gives us a sensible interpretation of the intercept by fitting Y around the sample mean of X.
example: if mean is 13.67 years of schooling then an individual at the mean is predicted to earn
$19.63.
Goodness of Fit correct answers The r2, the coefficient of determination, gives us an overall
measure of
the "goodness of fit" of the regression
◦ How well the regression line fits the actual Y values
Coefficient of determination correct answers R^2 predicts the overall measure of goodness of fit
R2 correct answers r2 measures the percentage of total variation in Y explained by the regression
model
TSS correct answers TSS = ESS + RSS
total sum of squares
What is econometrics? correct answers Tools of economic theory, math, and statistical inference
are applied to the analysis of economic phenomena.
How do we do econometrics? What are the steps involved? correct answers 1. Make a hypothesis
2. Collect data
3. Specify the model
4. Estimate the parameters
5. Check model specification - is the model good?
6. Test the hypothesis
7. Use the model to predict or forecast or answer questions
Simple linear model correct answers yi= B1 + B2Xi + ui
Bs are parameters/coefficients, X is regressor, u is error term
What is the error term? Why do we have an error term? correct answers omitted explanatory
variables,
aggregation of variables (consumption function)
model misspecification
may be determined by expectations rather than actual values
functional misspecification
x instead of x^2
measurement error
Linear correct answers linear in parameters:every term consists of a straightforward parameter
multiplied by a variable
linear in variables: every term consists of a straightforward variable multiplied by a parameter
when to reject the null correct answers reject: t-stat > t-critical
*don't forget to explain why: explanatory variable seen as significant at the % level
B2hat - t-crit*se(B2hat)< B2hat < B2hat+ t-crit*se(B2hat)
Method of OLS correct answers minimize the sum of the squared errors in order to have a very
fitted line
What is OLS? correct answers ordinary least squares
most popular application of regression analysis
What does OLS do? correct answers we sum the squared errors
Why do we use OLS? correct answers ◦ OLS is simple to do (compared to other regression
techniques)
◦ OLS has some nice properties - the regression line goes through the means
of X and Y and the sum of the residuals is zero (or very close to zero).
, ◦ OLS is the BEST estimator under a certain set of assumptions
How do we estimate the coefficients using OLS in a two variable model? correct answers
B2=sum(Xi-Xbar)(Yi-Ybar)/
sum(Xi-Xbar)^2
B1= Ybar - B2Xbar
- Be able to explain the intuition correct answers the goal is to minimize the sum of the squared
errors
Explanatory variable correct answers same as independent variable and regressor
Dependent variable correct answers Y or what is being explained by regressing data
Be able to draw a graph of a regression model and label all the parts (like the coefficients, the
error terms, etc) correct answers (look up in notes)
Causation and correlation correct answers Example:
A one dollar increase in income suggests a 51 cent increase in average
consumption
◦ Suggests or is correlated with (not causes)
Scaling and units and interpretation correct answers
Demeaning - what is it? Why do we do it? Interpret coefficient. correct answers We regress Y on
X* when X*=Xi-Xbar
gives us a sensible interpretation of the intercept by fitting Y around the sample mean of X.
example: if mean is 13.67 years of schooling then an individual at the mean is predicted to earn
$19.63.
Goodness of Fit correct answers The r2, the coefficient of determination, gives us an overall
measure of
the "goodness of fit" of the regression
◦ How well the regression line fits the actual Y values
Coefficient of determination correct answers R^2 predicts the overall measure of goodness of fit
R2 correct answers r2 measures the percentage of total variation in Y explained by the regression
model
TSS correct answers TSS = ESS + RSS
total sum of squares
