Human interest stories are an easy way for newspapers to fill space. They appeal to readers without having journalists spend much time with pesky fact checking. Especially as newspapers and other organizations lay off science journalists, it is increasingly clear that entertaining consumers, rather than getting the facts right, is what is important.
A good example is the case of Oscar the cat. In 2007, David Dosa, an assistant professor at the Alpert Medical School, wrote an essay in the New England Journal of Medicine describing a day in the life of Oscar. He and the staff of Steere House, a Providence nursing home where the cat lives, believe that Oscar has the ability to predict the deaths of patients. News agencies worldwide picked up this story in 2007 without considering whether there was scientific validity to it. The articles might have contained some token skepticism in an attempt to show balance — the false proxy of objectivity. They did not.
The Herald now has its own article on Oscar, which repeats the unfounded claim as credulously as anything from 2007. The piece, "Prof's book tells the story of a cat's eerie sixth sense" (Feb. 2), tells the story of Oscar and of Dosa's newly released book. Dosa's claims of Oscar's abilities are taken as fact and uncritically examined. After all, "Dosa published an article about him in the New England Journal of Medicine," so it must be true.
It turns out that the "article" is actually an essay, a narrative of a day when Oscar spent time by a dying patient. The important difference is that an essay presents no evidence to support its claim. The physicians and staff of the nursing home believe that their cat possesses the ability to sense when a patient is near death, but that does not mean that it is so. All we are given is an anecdote.
Many of the 2007 articles contain commentary on why a cat would be inclined to stay near a person just before his or her time of death. This, however, is premature. Before even bothering to speculate on the cause of Oscar's ability, we should first know whether the effect is real or whether the staff of Steere House is mistaken. Anecdotes might be useful for indicating new areas of research, but they are worthless for actually determining whether an effect exists. In this case, the staff's observations and Dosa's essay are cause for looking into this further, but they cannot be used to support the claim that Oscar has any unusual abilities.
I suspect that an actual study would find that Oscar cannot predict deaths, as this looks like a typical case of confirmation bias. Confirmation bias is caused by the tendency to remember evidence that confirms one's hypothesis while forgetting evidence that disconfirms it.
One of the classic examples of how confirmation bias works is the myth that emergency rooms get more visits during full moons. This results from emergency room staff experiencing a busy night and then discovering that the moon is full. Because this is an interesting coincidence, they tend to remember it more than other nights throughout the month that were just as busy. When staffers think back to when they were the busiest, they are more likely to say that it was around the time of the full moon. But when systematic studies are done, relying on records rather than fallible human memory, there is no correlation with the moon's phase.
The same effect has likely caused the staff of Steere House to believe that Oscar has abilities that he does not actually possess. Dosa has yet to publish a study showing the amount of time that Oscar spends with near-death patients as compared to other patients.
Dosa's essay claims that Oscar visits numerous patients throughout the day. It is therefore unsurprising that Oscar would visit a dying patient at some point before death. The staff remembers those visits while they forget the visits to patients who continue to live. Supposedly, the difference between those uneventful visits and the ones where Oscar predicts the patient's death is the length of time that Oscar spends with the patient.
However, Dosa gives no definite length of time at which Oscar's visit turns from friendly to grim. Nor is there a fixed amount of time between when Oscar stays with the soon-to-be-dead patient and when the actual death occurs. This means that it is easy for staffers to retroactively fit Oscar's behavior with what they expect to happen — perfect fodder for confirmation bias.
There is some good speculation on how Oscar could be predicting deaths, but without a systematic study to demonstrate Oscar's ability, the frailty of human memory provides a much better explanation for what is happening in Steere House. Until such a study is done, it is irresponsible for articles to propagate unfounded stories like those of Oscar the cat — even if only to entertain.
David Sheffield '11 is a math-physics concentrator from New Jersey. He can be reached at david_sheffield (at) brown.edu.
