I am trying to count the total number of households in a my dataset that have a child age 5 - 17. After restricting the data to the desired geography and qualifying child present nchild>0 & eldch>4 and yngch<18. I am trying to get the total of households in the data. After applying the person weight I have 499,123 individuals in that population, which makes sense. But when I try to count the total households by taking pernum=1 and applying the household weight I get 264,186 households. This only averages to 1.89 persons per household. How can this be given that all observations are based on individuals with children in the household. Shouldn’t I expect a number closer to 200,000?
Thank you for any help.