Originally Posted by
bblair
I know that there is a push by some organizations to put some light on this, but can't remember the name. For example, there are never negative studies published.
Compounding the problem is that the popular press sees these "studies"/press releases and publishes them with headlines, "New studies show....."
For example, the conventional wisdom, for whatever that is worth (zero) is that caffeine is an aggrevating factor in dementia and Alzheimers. Yesterday I saw a headline, "studies show..." and it was the exact opposite.
I am no where near an expert, but have read a lot of studies through the years (retired dentist) to know how to read the lit. The average schmo has no frikkin idea and the "journalists" who report it are even worse.
Maybe refering to
The Reproducibility Project ?
2 million studies a year... 1million can't be reproduced and there are only 400 retractions a year! 😳
They say "only" half is just outright fraud.. fake results... there are account of researchers taking Photoshop to their scans (as in a famous Alziemers research fraud) etc.. other reason are just bad research design.. bad or wrong math and stats analysis... something is very wrong...
And as you allude to, journals publish papers now even before they are even peer reviewed.. and press takes it up and runs with it.. even if they get the results right (quite often get wrong) half the time it will actually be wrong 😳
And on top of that, apparently there are also serious problems with the peer review process as well.. so you have bad/wrong research results out there for years.. that is used as basis for new research that 50% can also be wrong.. and on and on.. this is NUTS!!!
The analysis on cancer reseach in particular is that only 11% if it is reproducible!!!!! 😮