Tuesday, April 05, 2005
Crappy PC Research – or – Let Me Do A Study
Man I hate research that attempts to validate some (preconceived) PC assumption, but whose results rely on so many confounding variables as to make the study useless.
Here is one such study reported in CNN.com
“Study links TV viewing and bullying”
or from the article headline:
"Study: Kids who watch TV more likely to bully"
To quote the article:
“…At the same time, children whose parents read to them, take them on outings and just generally pay attention to them are less likely to become bullies, said the report from the University of Washington.”
This is typical confusion of cause and effect. How do people get money for these studies??? I’m not saying excessive TV watching is good, but this study clearly implies that watching TV is directly correlated with Bullying and is the causative factor. How about: “Are neglected kids more likely to watch TV?” Could I get grant money to study this question?
What makes this all the more infuriating is the implied: “something must be done or else…” Again, I’m not saying excessive TV watching is good, but this isn’t research, it’s advocacy. These people clearly had an agenda then went out to gather statistics that would induce action, or at a minimum garner more funds for further research into the “crisis”.
If your kid is a bully lets place the fault with poor upbringing, poor environment, and possibly random genetic factors. Good parents can sometimes have bad kids -- bad parents more often will -- growing up in a bad environment just about cinches the deal. TV is a minor contributor compared to afore mentioned factors. Of course this is just my opinion, maybe someone will fund me to do a study.
Links to this post:
1 Comments:
I have the same complaint with the "One glass of red wine a day reduces the risk of heart attacks" studies. Do any of them show causation? It seems to me that the same sort of person who is able to drink just ONE glass of red wine a day is also likely to have the sort of personality that would reduce the incidence of heart attacks. Correlation, NOT causality. Of course, I could be wrong.
By Anonymous, at April 11, 2005
