Recent reports that Oscar the cat can predict the impending death of hospice patients have been spread by the media rather uncritically (although my fellow science bloggers have already filled the skeptical gap left by the mainstream media). I was not going to write about it myself, but then yesterday “Freddy the Pig” left a comment in response to my seizure discussion asking about the so-called “seizure dogs” who can allegedly predict when their master will have a seizure. So I thought I would kill two birds with one stone.

Both cases bring up a few principles I want to touch on. The first is that before we waste too much time trying to explain the nature of a phenomenon we should first confirm that the phenomenon exists. This is a chronic problem in the paranormal research community, who spend far too much time weaving explanations out of quantum mechanics for phenomena, like ESP, that probably don’t exist. Admittedly, these can be useful thought experiments to assess the plausibility of a claim, but we need to resist putting the cart before the horse.

In both of these cases – Oscar the death predicting cat and seizure predicting dogs, neither has been scientifically validated. Both are likely a result of confirmation bias – the tendency to make and remember observations that confirm a belief we already have. For example, given that Oscar probably likes to lay next to warm people on their comfy beds, and these particular people are all very near death, it is likely that on occasion Oscar till spend time with someone right before they die. This, however, might seem rather creepy to nursing home staff, and rumors of Oscar’s apparent morbid ability would spread quickly. Now, everyone would be on the lookout for correlations between Oscar’s visitations and later death – to confirm the story. No one will much notice, report, or remember the times Oscar visited a patient and they did not die soon after.

Likewise, the notion of seizure dogs first emerged from people with epilepsy who had dogs as pets and noticed that their dogs acted differently right before they had a seizure. The same process of confirmation bias would then seem to confirm these initial observations.

It is very important to understand the power of confirmation bias – all of us fall victim to it on a regular basis. It is a ubiquitous cognitive defect in human reasoning. The antidote is controlled observation – in other words, science. Making observations in a systematic or controlled fashion eliminates bias as a factor.

In the case of Oscar the cat, we have no systematic observations – only highly dubious uncontrolled observation. The same is true of seizure dogs – there are no good controlled studies documenting a genuine ability to sense the onset of seizures.

Always with regard to such issues I hear from individuals who absolutely insist that the phenomenon is real because they have personally experienced it. But I find such personal stories completely unconvincing – confirmation bias is enough to account for it. Interestingly, another defect in human reasoning is that we intuitively find such personal stories very compelling – enough so even to trump scientific data. But history has shown that we must realign our intuition and trust controlled data more and uncontrolled experience less – even our own experiences.

One final point with regard to the Oscar the cat story – this was originally published in the New England Journal of Medicine. The NEJM is a highly prestigious journal with an excellent peer-review process and generally publishes high quality papers. I feel this was an error in judgment on their part. It is likely that they felt this was little more than a curiosity, a cute human interest story.

However, by printing the story they lend a great deal of credibility to the implicit claim that Oscar’s abilities are genuine. In so doing the NEJM has contributed in a small but real way to the scientific illiteracy of the public. The story reinforces the false notion that uncontrolled observations can be reliable and that seeming fantastical, even paranormal, powers are taken seriously by scientists. They missed the opportunity to teach the public about the pitfalls of bias in uncontrolled observation and the need for scientific study before concluding a phenomenon is real.

I don’t really expect this of every local newspaper relating the story (although that would be nice) but the NEJM should know better. I think this reflects a lack of appreciation on the part of mainstream scientific institutions for the problems of scientific illiteracy and pseudoscience in the general public and their responsibility in combating it (or at least not exacerbating it).