capture log close
set more off
log using basic_descriptives, text replace


/* an example of a basic descriptive analysis.

by modifying this basic syntax you can go pretty far in describing relationships
in your data.

Don't make the mistake of skipping the descriptive analysis in the rush to run a 
multivariate model.  You may miss something important.

The goal of the descriptive analysis is to so how close to the "final" answer you can 
get just using easy to understand methods.

y is your dependent variable
x w z q independent variables of interest
z and q are categorical, x and w are continuous

basic syntax

there are many more combinations of commands, but this is enough to get started.
*/

* generate data (you can ignore this part)
/*
clear
set obs 1000
corxz .5 x q2
gen w=int(30*(uniform()))
gen e1=invnorm(uniform())
gen e2=invnorm(uniform())
sort e1
gen z=group(2)
sort q2
gen q=group(5)

gen y=5*x+3*w+2*z+5*e2
replace y=y+10 if q==3
save basic, replace

*/
* begin syntax

clear
use basic

sum y x w z q

cor y x w z q

tab q, sum(y)
tab q, sum(y) nost nofreq

tab q z, sum(y)
tab q z, sum(y) nost nofreq

scatter y x
* see help scatter

sum w, det
gen w2=recode(w, 0, 5, 10, 15, 20)
* help functions

tab w2 z, sum(y)

egen miss=rowmiss(y x w z q)
tab miss
* help egen

tab q, sum(x)
egen avg_x=mean(x), by(q)

reg y x avg_x
sort q
by q: sum y x w z q
by q: reg y x z 



log close