7,455 reputation
1033
bio website
location San Jose, CA
age 34
visits member for 2 years, 10 months
seen 4 hours ago

Mad data scientist.


13h
comment PE-% Meaning in Statistics
If you can't cite the paper, how about the field? Finance, biology, cryptozoology?
14h
comment PE-% Meaning in Statistics
I think this needs more context. For instance, PE has a particular meaning in finance.
1d
revised Relationship between the latent variable as a function of regressors and the logit model?
added 7 characters in body
2d
comment Proxy variable approach to deal with endogeneity due to reverse causality
Criminologists have tried using the number of DUI and disorderly conduct arrests made per police officer as a proxy for proactive policing intensity in regression where more serious crimes were the outcome. I don't think this identifications strategy would placate most economists these days.
2d
comment Relationship between the latent variable as a function of regressors and the logit model?
Both are used by applied economists. In practice, the average marginal effects and predictions from the two are often quite similar, so the choice usually does not matter very much, except in cases where a lot of observations have probabilities near zero or one.
2d
awarded  Nice Answer
Aug
16
revised Relationship between the latent variable as a function of regressors and the logit model?
added 3 characters in body
Aug
15
revised Relationship between the latent variable as a function of regressors and the logit model?
deleted 28 characters in body
Aug
15
revised Relationship between the latent variable as a function of regressors and the logit model?
added 116 characters in body
Aug
15
answered Relationship between the latent variable as a function of regressors and the logit model?
Aug
15
comment Relationship between the latent variable as a function of regressors and the logit model?
Are familiar with utility from economics?
Aug
14
comment Determining dependancy of observations
What is the question you are trying to answer?
Aug
13
comment Stata error: “panels are not nested within clusters”
The clustering should not be on region-year pairs since, for example, the error for Bavaria in 2014 is presumably correlated with the error for Bavaria in 2013.
Aug
13
revised Stata error: “panels are not nested within clusters”
edited tags
Aug
13
revised Time dummies in panel data — absorbing effects?
deleted 21 characters in body; edited tags
Aug
12
comment Random Effects Tobit Model
Are all 94 districts in your data?
Aug
11
comment Which test to use to see if tumor weight increases at different stages?
@gung I don't believe nptrend requires panel data. You just need the groups to be ordered.
Aug
11
reviewed Approve suggested edit on Applied Analysis Question
Aug
11
comment Which test to use to see if tumor weight increases at different stages?
Isn't tumor size the outcome?
Aug
11
comment Which test to use to see if tumor weight increases at different stages?
Have you seen the UCLA test decision tree?