Tuesday, November 27, 2007

Statistics and Research

I have always been facinated with statistics and how easily they can be manipulated by a researcher to make a point. It is evident in athletic development or sport science research that in many instances the statistical analysis is very basic but still big claims are made from these papers. Looking through some research today there are many studies reporting "strong" corelations and than stating that X and Y are related with over 70% of the variance unexplained. To me I dont understand why editors allow researchers to report r= when it should be r squared. Maybe this is becaue the statistical knowledge of most people in the sports science field is poor. Be very careful when reviewing the claims of researchers and just because it is published is not a reson to believe. I know many people just see the title of a paper and than go to the conclusion nd take that concluding statement s gospel. Much of the problem may also stem from poor statistical power due to low cohort numbers and if there is low numbers than correlation should not be used. I dont want to go on too much but it is imperative to review the literature rather than just believe

No comments: