Vixen Van Debz
Well-Known Member
As a researcher and statisitician, this article drove me up the wall. Poor representation of results and poorly ran statistics are a bain in research, and the media make it worse by hamming up poorly done work. (See Gillian McKeith for an example of crap science gettings lots of positive press).
What the research says is "of the cheaters, most are dog owners, second are cat owners and third are horse owners". Well, given that dogs and cats are the most common of all pets in the whole population, then OF COURSE when looking at all the cheaters, more own dogs and cats and horses. I don't know if horses are the third mostly commonly owned 'pet', but I wouldn't be suprised if it's there or there abouts (maybe 4th if rodents are collapsed together). It also says that more men who own horses cheat than women who own horses. First of all, it doesn't say if this difference is statistically significant. Secondly, well conducted and analysed research has shown that men cheat more than women (if you consider that only the man is the cheater in a married man-mistress scenario). So, depsite advertising that essentially horse owners are amongst the greatest cheaters, what it actually seems to show is that because most people own dogs, cats or horses, that more cheaters are dog, cat or horse owners too. Whoopadeedo.
Another example is a study I saw saying smokers are significantly less likely to die of dementia than non-smokers according to research. Well yes, of course they do because they die too young to get Alzheimer's! If you control for age, then smoking and Alzheimers are in no way related. This work is guilty of similar over-simplifications and misunderstandings - it doesn't control for the disribution of these animals in the general population. *sigh*
Boring rant over!
What the research says is "of the cheaters, most are dog owners, second are cat owners and third are horse owners". Well, given that dogs and cats are the most common of all pets in the whole population, then OF COURSE when looking at all the cheaters, more own dogs and cats and horses. I don't know if horses are the third mostly commonly owned 'pet', but I wouldn't be suprised if it's there or there abouts (maybe 4th if rodents are collapsed together). It also says that more men who own horses cheat than women who own horses. First of all, it doesn't say if this difference is statistically significant. Secondly, well conducted and analysed research has shown that men cheat more than women (if you consider that only the man is the cheater in a married man-mistress scenario). So, depsite advertising that essentially horse owners are amongst the greatest cheaters, what it actually seems to show is that because most people own dogs, cats or horses, that more cheaters are dog, cat or horse owners too. Whoopadeedo.
Another example is a study I saw saying smokers are significantly less likely to die of dementia than non-smokers according to research. Well yes, of course they do because they die too young to get Alzheimer's! If you control for age, then smoking and Alzheimers are in no way related. This work is guilty of similar over-simplifications and misunderstandings - it doesn't control for the disribution of these animals in the general population. *sigh*
Boring rant over!
