Whoa, and I thought you understood. That statistic is the average of all paternity testing services. It's not a balanced test using statistical models to carefully randomly sample from each subset of the population in order to determine what percentage of the population are misattributed paternity cases, it's the statistics taken from everyone who has chosen to take a paternity test, which would mean every case where there is a dispute, suspicion, or reason to doubt paternity
only. So in other words, if you doubt the paternity of your own child enough to pay for paternity testing, based upon the overall statistics
among people who pay for paternity testing, you would have a 70% chance of being the father.
My hubby knows he's the father of my child, so he isn't going to have him paternity tested, so my hubby won't be included in those statistics. Do you understand? That statistic automatically
excludes all couples wherein the parents have no doubt about the paternity of the child!
Your example was flawed because unlike the sample group of "30% of everyone who ate at McDonald's" you aren't sampling everyone who had a baby, you are only sampling the people who chose to take a test.
Another example: If 25% of all diabetes tests come back positive, it doesn't mean that 25% of the general population has diabetes because only people who have reason to believe that they have diabetes get tested for it. The population of people tested for diabetes doesn't represent the general population.
Likewise, if divorced women have a high rate of misattributed paternity, it doesn't mean that women who are still married have the same rate of misattributed paternity. However, that article failed math, but what else would I expect from the Daily Fail?
"Its figures for 2007-2008, obtained under freedom of information rules, show that out of 3,474 DNA paternity tests ordered, 661 - 19 per cent - named the wrong man. It is the highest proportion since the agency started collating figures nationally."
According to that article, that percent refers to contested cases. The article doesn't state what percentage of all CSA mothers had their case contested. In another article it stated that
only 3% of all CSA cases are contested. 19% of 3% comes out to .6% of all CSA mothers with failed paternity tests. That's less than 1%. In fact, the second article states that sometimes multiple tests are ordered for the same mother, which would reduce that number even more. In the second article it states that currently 16% of all tests ordered fail.
From the article:
"There were 112,800 applications made to the Child Support Agency, with paternity contested in around 3 per cent of cases successfully so in 0.5 per cent of cases.
Read more:
Women could be forced to pay for paternity tests to prove identity of father's during CSA battles | Mail Online