In science, the most (in)famous dichotomy is created by p<0.05 and “positive” results indeed cannibalize “negative” results.

A recently published Registered Report reviewed the presence and reporting of null findings in the special education research base (authors searched all 2020 publications in 41 special education journals and identified 121 group-design intervention studies). The main result of their analysis was that only two (1.7%) articles reported all null findings and no articles reported null primary findings. 

But this paper received some publicity because of a new exploratory metric ‘p-values per study participant’. On average, there were 0.77 p-values reported for every study participant (SD = 1.35). The range was 0.02 (i.e., one p-value to every 50 participants) to 11.44 (i.e., more than 11 p-values to each participant!):