Hacker News Reader

Comment by deegles | Hacker News Reader

I've seen several references to the "replication crisis" in social sciences. Are there any prominent studies that have been replicated reliably? I'm just curious to look into them.

View on HN

In the largest of these replication studies, results of 39 of 100 studies were successfully reproduced. It's hard to wade through but all the info is here: https://osf.io/ezcuj/wiki/home/

Psychometrics? Although the results aren't always politically correct like some of the priming ones.

Actually priming, both with conscious amd subliminal cues is one of the most studied and replicated phenomena in cognitive neuroscience. Studied through behavioral and brain imagery data.

The Dunning-Kruger Effect is regularly replicated.

https://en.wikipedia.org/wiki/Dunning%E2%80%93Kruger_effect#...

The study you cite shows an opposite result.

Are there independent replications that actually show the same effect? I recall an attempted replication that produced equal miscalibrations on the upper and lower ends of the distribution, suggesting that the effect was just mean regression.

Your link also says that the Dunning-Kruger effect does not quite work for Asian cultures. I wonder if these persuasion tricks aren't also tied to culture. Do the tricks work in places/cultures were trust is generally low?

Trust is low in Asia, everyone is expected to lie as a general rule.

Reply on news.ycombinator.com