-repest- available on ssc

[Francesco Avvisati]

Thanks to Kit Baum, the programme repest is now available from SSC. To install it, type in Stata:

ssc install repest

repest estimates anything (as long as weights are allowed in the estimation of anything) using replicate weights (balanced repeated replication or brr weights, jackknife replicate weights,...), thus accounting for complex survey designs in the estimation of sampling variances. It is specially designed to be used with the PISA, PIAAC and TALIS datasets produced by the OECD. It also allows for analyses with multiply imputed variables (plausible values); where plausible values are used, the average estimator across plausible values is reported and the imputation error is added to the variance estimator.

Fans of the SSC moduels pisatools piaactools and pv may be particularly interested.

[Kristoffer Lander]

Dear Stata users.
My question relates to the new command - repest (ssc install repest), which is developed for OECD survey data sets. How this procedure calculates the standard errors? Are the standard errors robust when I estimate a panel of countries? And if I use pooled data (i.e. no variable which specifies the country id), are the standard errors clustered by country automatically?

[Luis Muelle]

Dear friends,
repest is a useful command for PISA data analysis. I would like to know how can I run a two or three multilevel model including school identification variable (schoolid).
In the help file I can not find an exemple with the appropiate sintax.
Can anyone help me on this?
Thanks in advance