Unpopular opinion: when you read a science paper, look at the graphs, and ask yourself “does this look like an actual effect? Does it pop out at me?” Most p=0.05 results look like “meh, basically the same, clusters overlap a lot” when displayed visually.
-
-
How are people going to get their pub count up if we start making them only report results that are both useful and reproducible?
-
I'm not making anyone do anything, just trying to allocate my attention better. I find that framing it as "bad scientists are self-interestedly pursuing career success And That's Terrible" trains the wrong mental habit (it shames wanting things)
- 1 more reply
New conversation -
-
-
Yeah. Wish one other # was standard to publish along with p: effect size normalized to volatility of the test subjects. Basically “how big was it and how big are things normally?”
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.