John Clees wrote:
Themis wrote:
if you cant validate the rawdata and interpret the numbers yourself
meaning I should personally interview the millions of american that got diabetes before I take it as truth?
no you dont have to do the interviews, that part is often made without manipulation.
but interpreting the data after the interview and drawing conclusions, thats where people get creative.
usually you read: diabetes rate 3 times higher than 1995.
but it is nowhere mentioned that they lowered the threshold value for diabetes since 1995 two times, also the newspapers dont say what they even took for a treshold value to draw this conclusion.
if i say the treshold value for diabetes is 90 mg/dl bloodsugar, then i can declare more than half the population sick of diabetes. if i say the treshold value is 120 its only 30 percent ..
the raw data (interviews, surveys, research) are mostly accurate.
but from raw data i could draw the wildest conclusions.
there is this saying
The only statistics you can trust are those you falsified yourself