Sent at the request of Sean Reardon
stata users of the -seg- command:
i have found a small bug in the -seg- command that i posted several weeks
ago. if the index option n was specified, by-groups with 0 counts in var1
or var2 were treated as missing and omitted from the results of the
command; note that this means that other index options specified as well
were not computed. a new version (version 2.01) that corrects this error
is now available. it can be downloaded using the ssc command (Stata 7) or
the archutil commands (Stata 6).
i have also added an option to the generate option in -seg- that allows you
to output a variable containing the number of units (schools, census
tracts) used to compute the segregation index for each by-group. see the
new help file for information on this option.
i do not subscribe to statalist, so if you have any questions about -seg-,
please contact me at email@example.com.