Can someone please help me with the Stata Commands for these question:

3. Define a number of cells (8 to 14) for the “Gert” variable and compute the conditional mean of “Kres”

for each “Gert” range. Plot the conditional means of “Kres” against the cell mid-point of “Gert” and join

the points. Explain the relationship of this graph to the population regression function.

I know what to do before the conditional mean... basically i don't know how to compute the conditional mean in Stata, and how do you plot the mean against the cell midpoint?

10. (a) Suppose that we are suspecting a relationship such as Gert , run a Park tests to determine if

there is heteroskedasticy in the model. Show your calculations and describe the procedure you are

following in detail. If we assume that kids is the source of the heteroskedasticity, compute a Goldfeld-

Quandt Test by omitting 4 observations from the center of the data (i.e. c=4).Comment. Compute aWhite

General Heteroskedasticty test and discuss the results.

I dont know the command to run a park test, and the Goldfeld-Quandt Test.

7. Generate 5 dummy variables for each year. Compare two regressions: (i) regress “Kres” on a constant

and four year dummy variables (omitting the first year dummy variable)(ii) regress “Kres” on five year

dummy variables and no constant. Interpret the results. Test the hypothesis that “Kres” has no yearly

component.

ok for this one... the data is for 5 years.... is this question asking to generate 5 dummy variables for each year... and since the data file contains over 20000 observations... how do you do generate a dummy variable for each of the years??

Thanks a lot!

3. Define a number of cells (8 to 14) for the “Gert” variable and compute the conditional mean of “Kres”

for each “Gert” range. Plot the conditional means of “Kres” against the cell mid-point of “Gert” and join

the points. Explain the relationship of this graph to the population regression function.

I know what to do before the conditional mean... basically i don't know how to compute the conditional mean in Stata, and how do you plot the mean against the cell midpoint?

10. (a) Suppose that we are suspecting a relationship such as Gert , run a Park tests to determine if

there is heteroskedasticy in the model. Show your calculations and describe the procedure you are

following in detail. If we assume that kids is the source of the heteroskedasticity, compute a Goldfeld-

Quandt Test by omitting 4 observations from the center of the data (i.e. c=4).Comment. Compute aWhite

General Heteroskedasticty test and discuss the results.

I dont know the command to run a park test, and the Goldfeld-Quandt Test.

7. Generate 5 dummy variables for each year. Compare two regressions: (i) regress “Kres” on a constant

and four year dummy variables (omitting the first year dummy variable)(ii) regress “Kres” on five year

dummy variables and no constant. Interpret the results. Test the hypothesis that “Kres” has no yearly

component.

ok for this one... the data is for 5 years.... is this question asking to generate 5 dummy variables for each year... and since the data file contains over 20000 observations... how do you do generate a dummy variable for each of the years??

Thanks a lot!

Last edited: