Bookmark and Share

Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

re:Re: Re: st: Direction of the effect of the cluster command on the

From   Christopher Baum <[email protected]>
To   <[email protected]>
Subject   re:Re: Re: st: Direction of the effect of the cluster command on the
Date   Thu, 6 Jan 2011 09:52:39 -0500

Austin wrote

Useful to think of super-obs but not quite right.  If you have 50
clusters and 100 regressors (with a few thousand obs) but you are only
interested in testing one coefficient, you will typically be fine,
i.e. you will have negligible bias in the SE thus getting correct
inference on average with the CRSE, and it may often be the case that
no alternative approach gets you correct inference (except resampling
clusters for a cluster-robust bootstrap).  So estimating a regression
with 50 obs and 100 coefficients is not quite the right analogy--more
useful to think of the "effective" sample size as between M (number of
clusters) and N (number of obs), computable using "roh" per Kish, L.
(1965), Survey Sampling, New York: Wiley (note that the CRSE is also
the standard svy estimator).

Quite so, Austin; unless you are interested in all the coefficients in a regression, you may not be that concerned about the number of 'super-observations'. The effective sample size is indeed a more useful construct.

However it should be noted, for those not that familiar with cluster-robust VCEs, that Stata uses the number of 'super-observations' minus 1 when it reports test statistics. For instance,

webuse grunfeld
reg invest mvalue kstock time, clu(company)

reports an ANOVA F-stat based on 3 and 9 df, where 9 is 10 companies - 1.  Likewise, the t-stat pvals are those for 9 df. This is rather important for the original poster of this thread, who was working with 4 clusters (and 3 denom. d.f. in the F, and 3 d.f. in the t).


Kit Baum   |   Boston College Economics & DIW Berlin   |
                              An Introduction to Stata Programming  |
   An Introduction to Modern Econometrics Using Stata  |

*   For searches and help try:

© Copyright 1996–2018 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   Site index