MarkW wrote:Open label studies are helpful to investigate where a problem may be found and help other investigators to look in the right places. Neuros will want to dismiss the study but they are not approaching CCSVI with open minds.
0.94 correlation is high but as the sample is small so the standard deviation is likely to be large. The detailed results will tell you this. (Ignore this comment unless you like statistics). To be honest I cannot be bothered to look up the SD on the full paper. All these correlation figures will vary until the testing methodology is refined and agreed, then training given.
Mark, I agree with you. What are your thoughts on how the results should be interpreted in such tests. If I am looking for a suspect element in a patient's body then generally positive is much stronger than negative.
For example, in the process of checking for an infection, a negative is never as strong as positive.
Don't know if I am communicating well here ...