Econ3951 Major Project Seminar

Text: Stock & Watson, "Introduction to Econometrics

1. I have data on the number of songs sung by each "Glee" cast member in the years 2009-2015. What type of data will this be?

(A) Cross-sectional data

(B) Time series data

(C) Panel data

(D) Categorical data

(E) None of the above.

2. Which of the following statements is true regarding instruments and controls?

(A) Controls are supposed to be related to the X variable in a regression, but instruments are not.

(B) Instruments are supposed to be related to the X variable in a regression, but controls are not.

(C) Controls are supposed to be related to the Y variable in a regression, but instruments are not.

(D) Instruments are supposed to be related to the Y variable in a regression, but controls are not.

(E) Neither instruments nor controls should be related to the Y variable in a regression.

3. In Lecture 3, I talked about a regression as a machine that turns certain inputs into certain outputs. Which of the following is not an input that goes into a regression?

(A) The specification of what type of regression to run

(B) The variables that appear in the regression equation

(C) The data that goes into the regression

(D) The estimated coefficients of the regression

(E) All of the above options are inputs that go into a regression.

4. Which of the following is not a categorical variable?

(A) zip code

(B) GDP

(C) whether or not a person passes the Capstone

(D) whether or not it snows in Minneapolis in April

5. Which of the following techniques could address the ceteris paribus problem in a regression?

(A) Obtaining a sample where X is set by a quasi-experiment.

(B) Adding control variables to your regression.

(C) Both of the above strategies will work to address the ceteris paribus problem.

(D) Neither of the above strategies will address the ceteris paribus problem.

6. You are interested in looking at the effect of education on earnings for a group of individuals. If you don't control for ability, which of the following is likely to happen as a result?

(A) The coefficient estimate will be more positive than the true value of the coefficient.

(B) The coefficient estimate will be less positive than the true value of the coefficient.

(C) The coefficient estimate will be more negative than the true value of the coefficient.

(D) The coefficient estimate will be less negative than the true value of the coefficient.

(E) Omitting ability from the regression will make no difference.

7. You run a regression of Y on two variables, X₁ and X₂. You find a p-value of .04 for X₁ and a p-value of .199 for X₂. Which of the below statements is definitely true?

(A) X₂ has a lower t-statistic than X₁.

(B) X₂ has zero effect on Y.

(C) The adjusted R-squared will be higher in a regression of Y on X₁ and X₂ than in a regression of Y on X₁.

(D) (A) and (B) are both true.

(E) None of the above are true.

8. The adjusted R-squared

(A) Tells you whether omitted variable bias is present in your regression.

(B) Tells you whether a given variable should be controlled for in your regression.

(C) Tells you whether a given coefficient is statistically significant.

(D) None of the above statements is true.

Questions 9 and 10 concern the following output table. We are looking into the impact of temperature on crop yields over the last ten years for a soybean farm in Minnesota. The following table gives the regression coefficient estimates.

Variable Crop Yield_t

Temperature_t -.154*** (.0349)

Temperature_t² .00108** (.000233)

Sunlight_t .0215*** (.00992)

Soil quality_t .667*** (.209)

Intercept_t 7.96*** (1.26)

N 100

R-squared .367

9. This regression uses

(A) Cross-sectional data

(B) Time series data

(C) Panel data

(D) None of the above.

10. Suppose that the current temperature is 56 degrees Fahrenheit. What is the predicted (approximate) change in crop yields if the temperature increases by one unit?

(A) -.154

(B) .06048

(C) -.03304

(D) -5.24

(E) .536