Hacker News new | ask | show | jobs
by irremediable 3895 days ago
Agreed. Also, I've not read the paper yet, but I'll be very interested to see the effect size. Lots of sex differences neuroimaging papers rave about significance while ignoring the minuscule scale of effects.
1 comments

Significance == publication, prestige, advancement

Effect power == bury it in the results, give it a one line 'yeah, we noticed that' in the discussion, so the publication reveiwer can't accuse you of missing something.

Indeed. It's all too tempting -- I've often found myself overlooking it a bit in my own work. When appropriate, I try to summarise key numbers, significance tests and effect sizes all in one table to avoid that problem.