10) GLYCEMIC INDEX
f) Using Bad Statistics to Judge the Glycemic Index
Statistics can be a funny thing. There are typically three possible statistical outcomes from an experiment: 1, that the data indicates that there is a very high probability that the treatment is having an effect, or 2, the data indicates that you can't say anything about the data (typically its just not a large enough sample, the treatment wasn't differentiated enough or there were confounding factors which could not be factored out), or 3, there is a very high probability that the treatment definitely has no effect. Note that proving a treatment has no significant effect is very difficult and typically requires a huge number of samples.
Much of the opposition to the glycemic index makes the error of equating lack of evidence in any given experiment as evidence that there is no effect. The two are not the same. This is an extremely common statistical error. One study saying that there is a relationship is worth ten studies which don't have sufficient sample size to make a definitive statement (in this case six studies on the glycemic index have shown a significant positive relationship and only three studies could make no definitive statement!). A more balanced approach is to simply recognize that the low glycemic index diets work for some and don't work for others. Try them and see if your particular metabolism works with the concept.
Current Chapter: 10) GLYCEMIC INDEX
a) Introduction
b) Types of Common Carbohydrates
c) Glycemic Index and A1c
d) Glycemic Index and Losing Weight
e) Glycemic Index Controversy
f) Using Bad Statistics to Judge the Glycemic Index
g) The Occasional High Glycemic "Splurge"
h) Glycemic Load
© Copyright 2009 All Rights Reserved.