Fewer papers are reporting findings on the border of statistical significance, a potential marker of dodgy research practices

Read original article