The Current Population Survey (CPS) is a monthly survey of about 57,000 U.S. households conducted by the Bureau of the Census of the Bureau of Labor Statistics. The CPS is the primary source of information on the labor force characteristics of the U.S. population. Details can be found at www.census.gov/cps and dataferrett.census.gov.

We are using a sample of 500 individuals from the March 2009 CPS. Here are three models for hourly earnings.

log(earningsi ) = b0 + b1hoursi + ei

log(earningsi ) = b0 + b1hoursi + b2educationi + ei

log(earningsi ) = b0 + b1hoursi + b2educationi + b3weeki + b4agei + ei

Each observation i represents a person. You can load data for this homework in two ways.

==============Version 1================== Copy and paste the code below in your R studio to directly read data from web.

cps

==============Version 2================== Download data named "df_hw1.csv" in the Homework folder on Canvas and modify the following code to load data in your R studio. Set your working directory to the folder where the data is located.

Question 5 1 pts What is the value of hi in Model 1? Question 6 1 pts What is the value of the R2 in Model 1'? Question If 1 pts What is the value of b1 in Model 2? Question 8 1 pts The value of hi differs. between Models 1 and 2 because Model 2 has a higher R-squared Model 2 adds a new variable [education] that is correlated with hours and log salary,r C} There is no way to know If} Model 2 adds a new variable {education} that is uncorrelated with hours and log salarvr Question 9 1 pts What is the value of hi in Model 3? Question 10 1 pts Suppose Person A and Person B have the same values for all the variables on the right hand side in Model 3, except Person A is 10 vears older than Person B. 1What is the difference in predicted difference in log earnings between the two people? Question 11 1 pts Suppose Person A and Person B have the same values for all the 1:rariah-les on the right hand side in Model 3, except Person A worked 45 hours per week and Person B worked 40 hours per week. What is the difference in predicted difference in log earnings between the two people? Question 12 1 pts Suppose a person were to get four more years of education. Would you expect them to earn higher salaries? Why? O Yes, because the coefficient on education in the regression is positive, which means that salaries tend to increase when people get more education, holding constant age, hours worked per week and weeks worked per year. O No, because the coefficient on education in the regression is positive, which means that salaries tend to decrease when people get more education, holding constant age, hours worked per week and weeks worked per year. O Yes, because the R-squared in Model 3 is higher than in Models 1 and 2 O The coefficient on education in the regression is positive, which means that salaries tend to be higher for people with more education, holding constant age, hours worked per week and weeks worked per year. However, increasing education may not increase salaries because correlation is not the same as causation.Question 13 1 pts Re-estimate Model 1, except add region as an additional right hand side variable. What is wrong with this regression? O The region variable is not cardinal. It indicates region, but its value (1, 2, 3, or 4) means nothing. O It excludes the education variable O The value of b1 is different from that in Model 1 O Its R-squared is lower than Model 3D Question 14 1 pts Consider the following two regression equations: hours; = do + an education; + vi log(earnings;) = bo + biVi + ei Estimate b1. This is a two-step process. First, you need to estimate the first regression model and save the errors. Then, you regress log(earnings;) on the errors (v;). What is your value of b1? D Question 15 1 pts Your value of bj in the previous question is O larger than in Model 1 O the same as in Model 2 O larger than in Model 2 O the same as Model 1