Practice exam
Name: ______________________________ Print your name on BACK (5 points penalty for writing on front!) Marketing and Price Analysis (AREC 403) Spring 2018: Makeup Midterm 2 Exam (125 + 10 Bonus points) 1. (100 points) Use the regression results from BRFSS (Behavioral Risk Factor Surveillance System) below to answer the questions that follow. Dependent Variable: Number of servings of vegetables per month (mean = 202.6) Variable Coef. t-stat. p- value Coef. t-stat. p- value Definition Mean Median Intercept 337.3 75.62 0.000 406.6 14.79 0.000 AGE -1.4 -1.32 0.187 Age of respondent in years 58.0 60 age_sq 0.0 1.42 0.157 Age of respondent in years, squared 3,642.4 3,600 Divorced -38.5 -4.82 0.000 = 1 if divorced; 0 = otherwise 0.162 0 Male -58.3 -9.89 0.000 = 1 if male; 0 = otherwise 0.411 0 Unemployed -36.0 -2.91 0.004 = 1 if unemployed; 0 = otherwise 0.060 0 Retired -17.8 -2.16 0.031 = 1 if retired; 0 = otherwise 0.363 0 Income 0.00051 4.59 0.000 0.00049 4.48 0.003 Median annual household income, $ 30,331.4 36,2119 R2 0.0029 0.0213 n = 7,371 a. (10 points) Explain very briefly the data generating process for this sample. b. (10 points) Does the regression coefficient of 0.00051 on income in the first regression have a causal interpretation? c. (10 points) What are the units of measure of the Male coefficient (-58.3)? AREC 403, Spring 2018 Midterm 2, Page 2 of 5 d. (10 points) Give a rough sketch of what the distribution of the divorced variable looks like? e. (10 points) Are the six additional variables in the second regression,taken as a whole, statistically significant? If you calculate anything, be sure to show all your work. f. (10 points) Calculate the marginal effect of age on the number of servings of vegetables. Be sure to show all your work and explain as concisely as possible what the marginal effect tells you. g. (10 points) Calculate the income elasticity of vegetable servings. Show all your work and justify the choice variables you use in the calculation. In plain English, write no more than two complete sentences explaining what the income elasticity tells h. (10 points) Some people think that getting workers back in the labor force after being unemployed has many good effects on workers besides just giving them a job. What impact would be the impact of eliminating unemployment be according to the model? AREC 403, Spring 2018 Midterm 2, Page 3 of 5 i. (10 points) Interpret the coefficient of the Retired variable, being sure the mention the units of measure. j. (10 points) Using the second regression results with seven explanatory variables, how many servings of vegetables would a person with median values of the explanatory variables be expected to consume? Write the equation you would use to calculate the expected water usage. AREC 403, Spring 2018 Midterm 2, Page 4 of 5 2. Multiple Choice (5 x 5 = 25 points total) Circle all correct answers for each question. Using a data set relating grades (measured in percentage) in AREC403 with absent dummy (= 1 if absent the day data were collected in class) by students, Thompson obtained the following results. Gradesi = 0.775 – 0.18 absenti , p-value of absent coef. is 0.004 & Correlation(Gradesi, absenti) = – 0.35 1. The regression results support Thompson’s hypothesis that students who were absent have, on average, lower grades. 2. Thompson should have received a grade of D in AREC403. 3. Students who were absent will have 0.35 percentage lower grade than those who were in class. 4. All of the above. Understanding the data generating process is important for: 1. gauging the quality of the data. 2. detecting measurement error. 3. cleaning those variables with missing values. 4. distinguishing correlation and causation 5. All of the above Including control variables is important because 1. having more variables will increase the R2 2. help isolated causal effects 3. They remove confounding effects on the coefficient of the explanatory variables of interest 4. they provide larger sample sizes for hypothesis testing 5. All the above. The R2 from any regression 1. measures how much variation in the dependent variable is accounted for by all the significant explanatory variables. 2. can take any value between -1 and 1. 3. may be used in some hypothesis tests 4. can usually be interpreted as a measure of causal effects. 5. All of the above. The income elasticity 1. is always negative. 2. is constant over different income level. 3. measures how the income is affected by the quantity consumed. 4. indicates the proportional change of income on the quantity consumed of that good. 5. None of the above AREC 403, Spring 2018 Midterm 2, Page 5 of 5 Bonus question (10 points) In an attempt to replicate Orley Ashenfelter’s results regarding the real rate of return on Bordeaux wines, a student unwittingly1 regressed the log of the price relative on winter rainfall measured in inches and winter rainfall measured in millimeters. Graph what the regression results would have looked like (Hint: it will be a plane-in-the-cloud graph). 1 Unwittingly means “without being aware; unintentionally.”