> Publish negative results. Or just don't publish the results at all, only publish the methodology and make the results available elsewhere along with the data

> Publish negative results. Or just don't publish the results at all, only publish the methodology and make the results available elsewhere along with the data and source code.