Censored Least Absolute Deviation model

[Byron Zamaasiya]

I am looking for a STATA syntax on CLAD models which can help me get p-values. I used this syntax (clad loffinc lsize urban coastal, ll(0) reps(200) ) and the results have no p-values

[Joao Santos Silva]

I guess you need to bootstrap it.

Best wishes,

Joao

[Byron Zamaasiya]

Thank you for your advice. Someone suggested that l adapt the following syntax ........permute y, reps(1000): clad y x1 x2 x3, reps(1). After running this syntax in stata l am getting the following result...expression list required. How can l address this new challenge

[Rich Goldstein]

as the message is telling you, you have not obeyed the instructions in the help file; here is an example from the help file, with the "expression list" (in this case, just one element) bolded

permute smoke x2=e(chi2), reps(10000) nodots: logit low smoke

[Rich Goldstein]

Byron wrote to me outside the Forum - please do not do that; if it does not say not to do that in the FAQ (I looked but could not find anything relevant), it should

[Byron Zamaasiya]

Apologies Sir. I tried what you suggested and still ran into challenges

[Rich Goldstein]

again, you don't show what you tried so no comment is possible

also, I am unclear on what you want -- if you need information on multiple variables, then I agree with Joao and you should use -bootstrap-

[Byron Zamaasiya]

Here is what l did. my syntax is as follows: clad index B1_Genderhh B2_Agehh B9_HHeduhigh B15_HHsize Visitextn FSS mktinfo FamilyRemit LivestockIndex,reps(1). With this syntax l get observed, bias, std error and confidence intervals. Now my issue is if l want to get regression results with p-values, how do l restate the above syntax. In an example that l saw, one of the dependent variables had an expression. for instance syntax 1: clad lwage exp school, reps(500) psu(cluster) and syntax 2: clad lwage exp school if rural==1, reps(500) psu(cluster) ll(1)
level(90) dots saving(c:\stata\data\boot1)