i see what you are trying to say and it makes theoretical sense, however, this is not a survey on the complete population. this was online poll on a sample of ~4400 people.
there would have to be impossibly huge number of siblings in that random sample so the difference could be explained that way.
say we simplify the sample to 2200 parents and 2200 kids. 220 parents (10% will say they are estranged from their kid).
estrangement is symmetric relation, so you should get the same answer from ~220 kids, but the poll got that answer from 220*1.6 = 352 kids. that would be 132 kids having sibling in that group of 220. in a random sample of 4400, that is just not possible (if the sample is truly random).
some parents are so delusional they don’t even realize their kids are not talking to them is much more logical explanation.
No, and the sample size isn’t terribly relevant. Yes, you could create a sample for which your conclusion was true, but it would hardly be the only possible set, or even the most likely.
As for symmetry, you seem to be ignoring the fact that for each mother or father, the relationship between parent and child is (potentially) one-to-many. Each parent can have many children. Therefore, many children can report estrangement to a single parent. This is one reason the numbers don’t match. Of course, things could go the other way, for example, where one child is estranged from two parents.
The relationship is not as symmetric as you suggest.
i see what you are trying to say and it makes theoretical sense, however, this is not a survey on the complete population. this was online poll on a sample of ~4400 people.
there would have to be impossibly huge number of siblings in that random sample so the difference could be explained that way.
say we simplify the sample to 2200 parents and 2200 kids. 220 parents (10% will say they are estranged from their kid).
estrangement is symmetric relation, so you should get the same answer from ~220 kids, but the poll got that answer from 220*1.6 = 352 kids. that would be 132 kids having sibling in that group of 220. in a random sample of 4400, that is just not possible (if the sample is truly random).
some parents are so delusional they don’t even realize their kids are not talking to them is much more logical explanation.
No, and the sample size isn’t terribly relevant. Yes, you could create a sample for which your conclusion was true, but it would hardly be the only possible set, or even the most likely.
As for symmetry, you seem to be ignoring the fact that for each mother or father, the relationship between parent and child is (potentially) one-to-many. Each parent can have many children. Therefore, many children can report estrangement to a single parent. This is one reason the numbers don’t match. Of course, things could go the other way, for example, where one child is estranged from two parents.
The relationship is not as symmetric as you suggest.
well, we don’t have details to reach a conclusion, so i conclude this debate ;)