I just learned about MHT. Have a question « Economics Job Market Rumors

3 days - 5 posts - Latest - RSS
Thread: 3 Goods vs 0 No Goods

Economics Job Market Rumors » Off Topic » Off Topic

I just learned about MHT. Have a question

Economist
2ef2

MHT as in Multiple Hypothesis Testing.
Is there any criteria for selecting a correction method? Or can I just Bonferroni everything.

3 days ago # QUOTE 0 Good 0 No Good !
Economist
94f9

reg y x

3 days ago # QUOTE 0 Good 0 No Good !
Economist
2ef2

Bump

3 days ago # QUOTE 0 Good 0 No Good !
Economist
d88b

Depends what group of tests you want to control the error rates for, and what concept of group error rate you want to control (e.g., familywise error rate versus false discovery rate). There is no right answer. The more conservative you are, the more conservative you are.

3 days ago # QUOTE 2 Good 0 No Good !
Economist
1a1e

Bonferroni is very underpowered if your p values are positively correlated.

3 days ago # QUOTE 1 Good 0 No Good !

Reply