Cluster-robust standard errors

[Tobjoh]

Dear PLS team,

I cannot see the possibility of using cluster-robust bootstrap inference (or jacknifing) in SmartPLS. Is there any way of adjusting the SEs one self to accomodate this or is it a feature that you coudl implement in the program in upcoming updates?

Many data sets include clustered data (time or sites) and just being able to control for this with fixed effects and not clustered SEs is a limitation for many users of SmartPLS.

Best regards,
Tobias Johansson

[admin]

If you point us to a reference that explains how this method works for bootstrapping, we may consider implementing this in future releases. Currently, it is not available. I would guess it involves some changes in the bootstrapping sampling process and not just changes in the calculation of the SE.

[Tobjoh]

It would be fantastic if such an option were available! It would incease the useability of smartPLS alot for many researchers. Especiallay for economics-based and sociological research, where this software is not that common. So, it is a promotion/marekting aspect for you to consider :-)


Yes, it involves some additional calculations I belive. STATA has implemented such approaches.

In Stata the pairs cluster bootstrap for OLS without fixed effects can be implemented in
several equivalent ways including: regress y x, vce(boot, cluster(id_clu)
reps(400) seed(10101)); xtreg y x, pa corr(ind) vce(boot, reps(400)
seed(10101)); and bootstrap, cluster(id_clu) reps(400) seed(10101)
: regress y x. The last variant can be used for estimation commands and user-written
programs that do not have a vce(boot) option. We recommend 400 bootstrap iterations for
published results and for replicability one should always set the seed.


Please see this text for more details:

http://cameron.econ.ucdavis.edu/researc ... bruary.pdf

Best regards
Tobias Johansson

[jmbecker]

Thank you for your suggestion. We will try to implement such a feature in future releases.