4 ms·
This is only 13%, so you could retract all of this and the only difference would be less "clinically confirmed" claims in commercials.
by foolmeonce 6y ago
This is only 13%, so you could retract all of this and the only difference would be less "clinically confirmed" claims in commercials.
- taeric 6y agoAre there no good results this would throw out?
- foolmeonce 6y agoFor every paper with an affiliation and a positive result create 4 copies, in the results put harmful or neutral instead of helpful. These are the missing papers according to the deviation between affiliated and unaffiliated. Now pick randomly from the papers you have.
- taeric 6y agoThat assumes you replace. More likely, for every paper you take out, nothing replaces it. Edit: that is, you are taking a zero sum approach.