I disagree with your use of "prove" when you wrote
Using "elaborate statistical techniques" one can prove pretty much whatever one wants to prove.
One may be able to massage data and create a nice looking chart to give an incorrect impression to unsuspecting people. This does not constitute a proof and it certainly won't pass muster with others who look at the same data and do their own statistical analyses.
Using "elaborate statistical techniques" one can prove pretty much whatever one wants to prove.
One may be able to massage data and create a nice looking chart to give an incorrect impression to unsuspecting people. This does not constitute a proof and it certainly won't pass muster with others who look at the same data and do their own statistical analyses.