The techniques described here have the special virtue of deriving an estimate for a hidden population while retaining respondent anonymity and the anonymity of network alters, but likely require larger sample size than the 132 personsinterviewed to attain
acceptable confidence levels for the estimate.