Talk:Wireless device radiation and health

edit

I just looked at the mineralization study that you mentioned It starts with " This study assessed differences in bone mineralization in the right and left hip of healthy male adult volunteers who were either nonusers of mobile phones (n = 24) or users who carried the phone close to the right hip, for at least 1 year (n = 24). Dual-energy x-ray absorptiometry (GE Lunar Prodigy) was performed in dual femur mode for each subject. Right and left hip bone mineral density (BMD) and bone mineral content (BMC) were compared. No difference in mean BMDs and BMCs between groups was found. " So at this point they have determined that phones have no effect. After that is where they do something statistically dodgy by breaking the results down into smaller subsets and nobody should be surprised that they find a correlation (although in a very small group). If you subdivide a group then you will find correlations just because you are looking at far more possibilities. If you regard a 1 in 20 likelihood as significant but look at 100 subsets then even for random numbers you would expect to find 5 correlations and there is no way to know how many ways they divided this data into subsets. Mtpaley (talk) 01:06, 12 April 2020 (UTC)Reply