Hello everyone,

I am doing an analysis on "Socio economic inequality on childhood health outcomes". I am considering stunting as my outcome variable. With the preliminary analysis I found that children belonging to lower wealth quintiles (25ht percent) are more stunted followed by 50th percentile children and so on.

stunted poorest poorer middle richer richest Total

0 61.32 66.67 73.52 80.72 83.41 74.73
1 38.68 33.33 26.48 19.28 16.59 25.27

Total 100.00 100.00 100.00 100.00 100.00 100.00

I am planning to do a quantile regression. I hope this regression helps e in understanding how the first 25 quantiles (i.e., first 25 percent sample with height for age scores), 50th quanilte and 90th quantile distribution have various independent variable effects.

As a first I have plotted stunted scores by using the following command in STATA

kdensity stunt_child, normal (the graph attached below).

I am planning to run a quantile regression, stunted_scores being my dependent variable. Residence, wealth, education f mother, age of child and other being my independent variables. So before directly running a quantile regression, what are the perquisite I have to follow with my independent variables. Do I need to check anything to include the variable in my models or else consider all the variables of the literature and run qreg.
One more, Before including these categorical variables in my model shall i create dummy variables for each independent variable and then do analysis ? or else STATA will take care of that.

Thanks in advance.