Not applicable
Posts: 1

# pairwise comparison of siblings

Hello

I have a dataset that includes a large number of siblings. I would like to do pairwise comparison of these siblings. More specific, I am going to analyze if siblings, who are discordant on e.g. marital status have an equal risk of being unemployed.
I can identify siblings, because the dataset includes information about the parents of each person in the dataset. With this knowledge - how can I prepare data for - and do the pairwise comparison of the siblings?

Kind regards Susan A
Frequent Contributor
Posts: 79

## Re: pairwise comparison of siblings

It would be easy to analyse the problem if you can post some sample data and write down what you want out of it.

Regards

Dhana
Allianz Cornhill Information Services.
Not applicable
Posts: 0

## Re: pairwise comparison of siblings

Okay - here is a sample of my data:

Observation......Variable:marital status.....Var:health.........Var:mother........
Person 1.................................1...........................3.............."mother code 1"
Person 2.................................1...........................3.............."mother code 1"
Person 3.................................2...........................1.............."mother code 1"
Person 4.................................1...........................2.............."mother code 2"
Person 5.................................1...........................3.............."mother code 2"
Person 6.................................1...........................3.............."mother code 3"
Person 7.................................2...........................4.............."mother code 3"

I am going to find out, if two siblings, who are discordant on e.g. marital status at age 30, experience the same kind health outcomes later in life. In order to find out which of the persons in my dataset are related, I am going to use the "mother code". If two (or more) persons have the same mother code, they are related. So in other words, I am going to perform a fixed effect analysis of each of the sibling pairs in my dataset. If a person has more than one sibling, the pairwise comparison will be made between all of them (as long as they are discordant on marital status).
I would like to start finding out the number of siblings who are discordant on marital status (and later on perform survival analysis). But I dont know how to arrange my data in order to do this pairwise comparison.