if you are using stata, the code can be summarized in 3 codes mentioned below with very brief notes;

.pca variable_x - variable_y //(this represents dichotomized variables, and continuous asset variables chosen consciously, replace variable_x-variable_y with a list of your assets)

. predict wealthscore //(by putting only one name; you get eigen vectors, or weights, of the first principal component, which we assume is reflective of wealth)

. xtile quintile= wealthscore, nq(5) //this generates quintiles

Great, thanks David Mukunya

Steps

1. pca variable1, variable2, variable3 (ensure all variables are binary coded 1 for yes (available) and 0 for No (absent))

2. predict wealth_indexscore

3. xtile wealth_index= wealth_index_score, nq(5) (IF YOU WANT FIVE EQUAL GROUPS)

Thanks Solomon Wafula,

David Mukunya Solomon Wafula so please i do not understand last step

also i have some qeastion

1- most standaration data befor starting PCA

2- Most i do id index

3- how about negative values

thanks so much