In the 2012 March CPS, using num of own children - nchild. The field value doesn't match the person records.

I’m getting odd results using the variable nchild in the march 2012 CPS. When I look at actual data records I find a household with famsize = 3. The head is 44, spouse is 39, and the third person is a parent age 65. nchild for the head and spouse = 0 but nchild for the parent record = 1. How can that be? This is only one example as I find many other inconsistencies.

I’m sorry, but I don’t think I understand your concern. Without looking at the specific family, it seems as though all three persons are related (famsize = 3), with the either the head or the spouse being the child of the parent. The nchild variable counts the number of own children in the household regardless of the child’s age, so this situation seems sensical to me. If you believe there is a problem with this family, or any other, feel free to reply with a household serial number so I can look at the same family you are looking at.