The Guardian today has an article about a research paper published back in May in the British Journal of Psychiatry. There are more and more attempts to control the future in our society. Predictive statistical models are increasingly being used by everyone from supermarkets (to “target” their marketing to you on the basis of what they think you might like to buy), to social work (to give special help to young mothers who they think might give birth to children who will become criminals), to the criminal justice system (to try and predict re-offender probability), to (my main area of interest) health care (where the experience of groups is used to determine what interventions an individual should or should not recieve – so called “Evidence Based Medicine”).
The paper discussed in today’s Guardian shows that the margin for error between the group studies and individual outcomes is so great that –
When applied to individuals the margins of error are so high as to render any results meaningless.
Almost every day I have a discussion with patients about risks and choices. I always emphasise that the statistical predictions are based on groups and averages and that there is absolutely no way of knowing to what extent they are relevant to this individual.
We are all different. Nobody, but nobody, can tell an individual what their future holds and to pretend they can on the basis of statistical modelling which isn’t up to the job is potentially very harmful.
This heroes not zombies site is about encouraging people to become aware, to think, and to develop their uniqueness. We need to celebrate individuality and difference more and we need to understand that people matter more than statistics – especially in social work, justice systems and health!
Agree with you. Thanks for sharing this article and your ideas about it.
[…] of “proven” or “unproven” treatments to be very unhelpful. Not only because no treatment will work for every patient, so a treatment is only “proven” for that person when we see how things turn out for […]
I dont get this……..what does the article have to do with the tiltle….??
[…] >> Almost every day I have a discussion with patients about risks and choices. I always emphasise that the statistical predictions are based on groups and averages and that there is absolutely no way of knowing to what extent they are relevant to this individual. We are all different. Nobody, but nobody, can tell an individual what their future holds and to pretend they can on the basis of statistical modelling which isn’t up to the job is potentially very harmful. Why your uniqueness matters […]