econ 382
NAME: ANSWER KEY
Econometrics First Test, Queens College, I. R. Kelly
Thursday, September 27th, 2012, 9:25-10:40am
Please put all answers on the question sheet. You have one hour and fifteen minutes to finish the test. Best of luck.
1. (6 points) How does a panel data set differ from a regular pooled data set? Give a specific example.
A pooled data set is one that combines both time series and cross-sectional data; in other words, there are many cross-sectional units over a period of time. A panel data set is a specific type of pooled data set in which the same cross-sectional units are followed over time. An example would be collecting data on the unemployment rate in
New York, New Jersey, and Connecticut from 1990 to 2000.
2. (4 points) You have estimated the following equation:
ˆ
Yi b1 b2 X i
where Y = Length of stay in office for senator i and X = Senator i’s popularity
What do you think the sign of b2 is (positive or negative)? Why? (Note that X is the independent variable and Y is the dependent variable.)
The sign of b2 is likely positive; the more popular a senator is (i.e., the more people there are that like his/her policies and ideas), the more likely people are to vote for and re-elect that senator, thus lengthening his/her stay in office. 3. (6 points) The auto.dta file (partially shown here) is a good example of what type of data set (i.e., is it cross-sectional, time series, pooled, panel)?
Why?
This file is a good example of a cross-sectional data set, since data are collected for many cross-sectional units (types of automobiles) at one point in time. 1
4. You obtain the following in Stata after summarizing the variables age and wage in detail:
. sum age, detail age in current year
------------------------------------------------------------Percentiles
Smallest
1%
34
34
5%
35
34
10%
35
34
Obs
2246
25%
36
34
Sum of Wgt.
2246
50%
75%
90%
95%
99%
39
42
44
44
45
Largest
45
Econometrics First Test, Queens College, I. R. Kelly
Thursday, September 27th, 2012, 9:25-10:40am
Please put all answers on the question sheet. You have one hour and fifteen minutes to finish the test. Best of luck.
1. (6 points) How does a panel data set differ from a regular pooled data set? Give a specific example.
A pooled data set is one that combines both time series and cross-sectional data; in other words, there are many cross-sectional units over a period of time. A panel data set is a specific type of pooled data set in which the same cross-sectional units are followed over time. An example would be collecting data on the unemployment rate in
New York, New Jersey, and Connecticut from 1990 to 2000.
2. (4 points) You have estimated the following equation:
ˆ
Yi b1 b2 X i
where Y = Length of stay in office for senator i and X = Senator i’s popularity
What do you think the sign of b2 is (positive or negative)? Why? (Note that X is the independent variable and Y is the dependent variable.)
The sign of b2 is likely positive; the more popular a senator is (i.e., the more people there are that like his/her policies and ideas), the more likely people are to vote for and re-elect that senator, thus lengthening his/her stay in office. 3. (6 points) The auto.dta file (partially shown here) is a good example of what type of data set (i.e., is it cross-sectional, time series, pooled, panel)?
Why?
This file is a good example of a cross-sectional data set, since data are collected for many cross-sectional units (types of automobiles) at one point in time. 1
4. You obtain the following in Stata after summarizing the variables age and wage in detail:
. sum age, detail age in current year
------------------------------------------------------------Percentiles
Smallest
1%
34
34
5%
35
34
10%
35
34
Obs
2246
25%
36
34
Sum of Wgt.
2246
50%
75%
90%
95%
99%
39
42
44
44
45
Largest
45