Dear Stata users,
I have a cross-sectional dataset where I test the effect of a county-level predictor "xvar" on a city-level outcome "yvar". Each county has several cities. The predictor values are the same for all cities in any given county, whereas the outcome varies by city (please see the example below). I wanted to ask how to generate another variable, say xvar2, where xvar values will be randomly assigned to different counties. So that, for example, all cities in Appling county end up with xvar=.32336795, those in Baldwin with var=.3913819 etc but the values assigned randomly. I am working with a cross-sectional dataset and I would like to reproduce a placebo test hoping the effect won't hold with random assignment. the fact that cities are nested into counties complicates the assignment for me. Ideally, I would like to run several regressions with different random assignments and tabulate the results with a neat line of code.
Thank you in advance for any help.
I have a cross-sectional dataset where I test the effect of a county-level predictor "xvar" on a city-level outcome "yvar". Each county has several cities. The predictor values are the same for all cities in any given county, whereas the outcome varies by city (please see the example below). I wanted to ask how to generate another variable, say xvar2, where xvar values will be randomly assigned to different counties. So that, for example, all cities in Appling county end up with xvar=.32336795, those in Baldwin with var=.3913819 etc but the values assigned randomly. I am working with a cross-sectional dataset and I would like to reproduce a placebo test hoping the effect won't hold with random assignment. the fact that cities are nested into counties complicates the assignment for me. Ideally, I would like to run several regressions with different random assignments and tabulate the results with a neat line of code.
Thank you in advance for any help.
Code:
* Example generated by -dataex-. To install: ssc install dataex clear input str8 county long id str22 city float(xvar yvar) "Appling" 1 "Baxley" .3488717 .77 "Appling" 1 "Graham" .3488717 .59 "Appling" 1 "Pine Grove" .3488717 .32 "Appling" 1 "Surrency" .3488717 .17 "Atkinson" 2 "Axson" .8689333 .81 "Atkinson" 2 "Pearson" .8689333 .31 "Atkinson" 2 "Willacoochee" .8689333 .22 "Bacon" 3 "Alma" .32336795 .72 "Bacon" 3 "Guysie" .32336795 .7 "Bacon" 3 "Rockingham" .32336795 .91 "Baker" 4 "Elmodel" .32336795 .68 "Baker" 4 "Newton" .32336795 .35 "Baldwin" 5 "Hardwick" .5844658 .74 "Baldwin" 5 "Milledgeville" .5844658 .19 "Banks" 6 "Baldwin" .3913819 .31 "Banks" 6 "Homer" .3913819 .14 "Banks" 6 "Maysville" .3913819 .65 end label values id id label def id 1 "Appling", modify label def id 2 "Atkinson", modify label def id 3 "Bacon", modify label def id 4 "Baker", modify label def id 5 "Baldwin", modify label def id 6 "Banks", modify
Comment