Stata Best Practices – Quiz 3

Quiz Question Explanations

Question 1

First Stata stores “10/5” in the local macro “num”. Stata does not execute the calculation because I did not include “=” when defining the macro; i.e. I typed “local num 10/5” not “local num = 10/5”. For the display command, Stata replaces the macro with the value “display 10/5”. Since “display” acts like a calculator when it’s given a numeric function, then Stata executes the calculation, and returns “2”.

Question 2

First Stata stores “10/5” in the local macro “num”. Stata does not execute the calculation because I did not include “=” when defining the macro; i.e. I typed “local num 10/5” not “local num = 10/5”. For the display command, Stata replaces the macro with the value “display “10/5″”. Since there are quotation marks around “10/5”, Stata treats it like a string and returns exactly what is in the quotation marks without attempting any calculation, i.e. “10/5”.

Question 3

Since I include “=” when I defined the local macro, Stata attempts to execute the calculation. Since 10/5 = 2, Stata saves the number 2 in the local macro num. Then when I display the value of this local macro, Stata returns 2. (Note that since I display num in quotation marks, Stata is treating “2” like a string rather than a number. If I had wanted to do any calculations with “`num’”, then I would have had to remove the quotation marks from `num’ in the display command).

Question 4

Stata stores the results of calculations in internal macros that follow the “r([macro])” naming convention. For some estimation commands (like regress), Stata stores results in “e([macro])”. You can view the stored macros after a command by typing “return list”. If you run a summarize command and then type “return list”, you’ll see that Stata stored the standard deviation value in the local macro “r(sd)”.

Question 5

Stata stores the results of calculations in internal macros that follow the “r([macro])” naming convention. For some estimation commands (like regress), Stata stores results in “e([macro])”. You can view the stored macros after a command by typing “return list”. If you run a t-test command and then type “return list”, you’ll see that Stata stored the p-value from a two-sided test in the local macro “r(p)”.

Question 6

To view the answer, run the following commands on the EG DIB baseline dataset:
sum hindi_bl, det
disp `r(p99)’

or you can type “return list” in the command line after running the summarize command.

Question 7

Hardcoding refers to manually entering values from previous calculations or functions rather than saving and using them programmatically. This can lead to copy-paste errors, errors from rounding (e.g. if Stata only displayed the first 10 significant digits, but the actual value has more significant digits), and may be inefficient in long code or code that you update and run multiple times.