1. This problem continues work with the California test scores data set. Suppose
you are thinking of adding some variables to the model that included
str, expn stu, and avginc as explanatory variables.
A variable of interest to you is the percentage of students in the district who
are from families eligible for Cal Works (calw pct). Add this variable and
interpret your results.
2. Now add two other variables of interest and interpret your results. (Continue
to include the added calw pct variable from the previous question.)
The variables are meal pct, the percentage of students eligible for reducedprice
meals, and el pct, the percentage of students who are “English learners”.
3. Comparing the results from the two model you have estimated, explain
the role played by the correlations between the three percentages you have
added in these models, in interpreting the results.
4. Data compiled by Professor Jeffrey Williams included grades earned by
Managerial Economics students in ARE 106, grades in 100A, and grades
in Statistics 103. He suggested the model below:
g106i = 1 + 2g103i + 3g100Ai + ui
to predict grades in 106. Use the data set 106grades.dta to estimate this
model and provide an interpretation of the results.
5. Professor Williams included an additional variable, TRANSFER, which equals
1 if the student in question was a transfer student and 0 otherwise. Repeat
the previous question for transfer students only and then for non-transfers
only; do you see any differences in the results that seem worth noting? Can
you provide a verbal explanation for these results?
6. Use the Consumer Expenditure Survey (CES) data set from Dougherty (under
Stata data sets on our Smartsite) to estimate the model below:
FDHOi = 1 + 2EXPi + 3SIZEi + ui