Is there a journal that publishes studies that are actually good, like the journal checks to make sure there's no glaring failures like p<.1 n=10 correlations tested=100 before publishing?
Conversation
Replying to
When I was on the editorial board of 'Intelligence', I think we were pretty good at quality control for publishing papers. Small niche journals with smart experts in a domain with a strong track record of empirical results tend to be pretty good at spotting nonsense.
4
21
Show replies
Replying to
I think most respectable journals meet that bar. Which those are though depends on the area. One problem for outsiders is there's so many low-quality journals it's hard to tell the good from the bad. Really, you need an area expert.
1
20
Show replies
Show replies
Replying to
No, scientific journals are cringe and anything free is slow or "illegal"
youtu.be/PriwCi6SzLo
Hears a neat video that's related tho.
Replying to
Depends on the field. In economics you can be sure that such failures don't occur in most top 5 journals. Especially Econometrica has high standards. But a field like medicine or psychology has no such journals afaik
3
Replying to
Referees at good journals should recommend to reject papers that have bad methods and hopefully they won't get published. You can check whether a journal is good roughly by how much its articles are cited.
1
Replying to
The review process in almost all *decent* journals will spot obvious statistical failures.
But clearing this bar does not guarantee in any way that the study is "actually good", nor protects against fabricated data, etc.
6
Replying to
It's not exactly what you asked for, but "Preclinical Reproducibility and Robustness" is likely to be of interest:
Replying to
"Journal of Comments and Replications in Economics" publishes replications (and failures to replicate) of economic studies:
jcr-econ.org/aims-scope/
1
Replying to
For natural/life sciences, usually the higher-IF *field-specific* journals are more reliable. E.g. in chemistry, I find JACS or an ACS sub-journal to be more reliable than Science/Nature and than MDPI and their ilk. Cell Systems more than Cell etc.
1
3