Well, not really a bad science blog post but it seems some people are willing to go to all manner of lengths to avoid facing up to statistical analysis they don’t agree with.
Lurking on scienceblogs once more, I came across a post titled More guns, more homicide in which Ted Lambert discusses a study showing that American states with the highest incidence of gun ownership also have the highest incidence of homicides. Yes, I know that seems to be so self evidently obvious there is an urge to go “duh” but studies like this are required to ensure that what we assume is the case, is the case.
Anyway, it is not Ted’s post (as such) nor the study itself I am talking about now but the inevitable furore from the pro-gun lobby.
Take this little chestnut from Jeff Stoyer:
In the current study, they claim they’ve “controlled” for factors such as unemployment, etc. I’d be interested in seeing how they accomplished that statistical dance.
Well blow me down. Who would have thought that any statistical survey would have been able to carry out this sort of “statistical dance.” It is mind boggling and sadly pays testament to how little people learn about statistics and the methods involved.
Matters are certainly not helped by the old adages: “89% of all statistics are made up on the spot,” “Lies, damn lies and statistics” and “you can make statistics prove anything you want.”
Sadly, all this points towards the “general public” thinking that statistics is some odd combination of witchcraft and subtle dishonesty. Now, I know that public opinion is not science but it does influence policy much more than science or statistics will ever manage. The more that people are swayed by a simple lack of understanding about how statistics are collected and the implications they produce, the more governmental policy will veer away from the track marked “sensible” and towards the one marked “Insane – Faith Schools – Creationism – Flawed Criminal Policy [etc].”
I will leave a rant about crime statistics and how they are warped and misinterpreted for another day and hopefully there are those who are more statistically inclined than myself who can add more to this topic.