You do not need a statistics degree to spot a weak paper. Four checks filter out most of the noise before you read a single result.
Check the sample and the design
Small, unblinded, single-site studies generate headlines and rarely replicate. Randomised and pre-registered work deserves more of your attention.
Read the effect size
Statistical significance says the effect probably is not zero. Effect size says whether it matters to anyone.
