Catherine Nixey looks at over-the-counter paternity tests:

Precise statistics on human infidelity are hard to come by. What evidence there is tends to indicate that human lovebirds are little better than their feathered counterparts. In 1970 a group of researchers looking into blood groups tested the blood types of inhabitants in a block of flats in Liverpool. They were startled to see that their results indicated a paternal discrepancy of 20-30%. Thinking, perhaps unfairly, that this might be something to do with Liverpudlians, they moved south and repeated the test, only to find similar results. In 1984 a group of scientists in Nottingham looked at women seeking fertility treatment because their husbands were sterile. Despite their husbands’ sterility, 23% of the women managed to become pregnant before receiving treatment.
Other studies have produced a more comforting picture. Recent research in Sweden and Iceland found rates of non-paternity between 1% and 2%. But while these figures may be reassuring in one sense, scientifically they are far from comforting. The disparity between them is enormous. Clearly large-scale, randomised testing is needed to find reliable average levels of non-paternity. The results would not just be interesting but useful in areas such as heritable diseases. There’s just one problem: such tests could be a source of considerable distress. As a result, much of the information that is available on paternity has emerged, like the 1970 Liverpool study, as a by-product of studies with other aims.

Polly Vernon took a magnifying glass to modern infidelity last month. The Babies trailer via Kottke.

We want to hear what you think about this article. Submit a letter to the editor or write to letters@theatlantic.com.