Hello All,
I've been posting across the Stata forum with regards to a dataset I'm working with in which the unit of analysis is directed-dyad year. I like to thank everyone that has helped me along the way, as this is the first time I have really delved into directed-dyads (country a-country b year, country b-country a year). The data is also time-series, cross-sectional (TSCS).
Around this time last year, someone posted a topic that didn't receive any answers regarding dyadic clustering here:
http://www.statalist.org/forums/foru...ering-in-stata
The Aronow article pretty much states that if dyadic clustering is not taken into account in models, it will fail to account the possibility of inter-dyad correlations. One of the ways to account for dyadic clustering is to include a Huber-White sandwich estimator, as dyads are often non-independent.
That being said, I am still trying to learn more about GEE (generalized estimating equation)models.
In the Stata manual regarding GEE under standard errors, it states:
So, does this mean that if I wish to cluster around dyads using GEE, I can't use
, but if I declare the panel to be dyads specifically,
, and use
it would be as if those it is clustered around the panel variable? I just want to make sure I read that correctly.
I've been posting across the Stata forum with regards to a dataset I'm working with in which the unit of analysis is directed-dyad year. I like to thank everyone that has helped me along the way, as this is the first time I have really delved into directed-dyads (country a-country b year, country b-country a year). The data is also time-series, cross-sectional (TSCS).
Around this time last year, someone posted a topic that didn't receive any answers regarding dyadic clustering here:
http://www.statalist.org/forums/foru...ering-in-stata
The Aronow article pretty much states that if dyadic clustering is not taken into account in models, it will fail to account the possibility of inter-dyad correlations. One of the ways to account for dyadic clustering is to include a Huber-White sandwich estimator, as dyads are often non-independent.
That being said, I am still trying to learn more about GEE (generalized estimating equation)models.
In the Stata manual regarding GEE under standard errors, it states:
vce(robust) specifies that the Huber/White/sandwich estimator of variance is to be used in place of the default conventional variance estimator (see Methods and formulas below). Use of this option causes xtgee to produce valid standard errors even if the correlations within group are not as hypothesized by the specified correlation structure. Under a noncanonical link, it does, however, require that the model correctly specifies the mean. The resulting standard errors are thus labeled “semirobust” instead of “robust” in this case. Although there is no vce(cluster clustvar) option, results are as if this option were included and you specified clustering on the panel variable.
Code:
vce(cluster clustvar)
Code:
xtset dyad year
Code:
vce(robust)
Comment