Question special

The purpose of research is to confirm or reject the null hypothesis but negative / non-significant results have been traditionally more difficult to publish in academia due to publication bias. However, research is, by its nature, a collaborative discipline, and both positive as well as negative results should all be significant even if statistically it is not so, as they help us achieve new direction towards a working theory. Any thoughts or words of advice from our panelists, co-moderators and followers on how to publish non-significant results?