standard error of proportions/clustering

[Deepa Bhat]

To the statisticians in the group:

Pardon me for asking this and only giving part of the information for
this.   But when someone writes this, what does this mean?

"I ran the SAS code to get standard errors of proportions assuming
clustering within districts (have to combine over sites) for the 2
districts in each intevention group.  The results are:
group 1 -- 0.14
group 2 -- 0.90
group 3 -- 1.28
Assuming clustering within districts only clearly isn't valid.

What does the last sentence mean? What does it mean that the standard
error of proportions for each group is different?

Thanks,
Deepa