. Author manuscript; available in PMC: 2016 Jan 31.

Published in final edited form as: Demography. 2015 Feb;52(1):329–354. doi: 10.1007/s13524-014-0358-x

Table 3.

Description of imputation methods

Method	Description
1 Logical	In target data, those appearing to be “probably legal” are coded as legal and the remaining as unauthorized.
2 Demographic Accounting	In target data, those appearing to be “probably legal” are coded as legal. Among remaining, single-imputation method used to assign targeted percentage as unauthorized. The rest are coded as legal.
3 Single Imputation	In target data, unauthorized = 1 if random draw from uniform distribution < predicted probability of being unauthorized; unauthorized = 0 otherwise; predicted probability is generated from a prediction equation estimated on the donor sample.
4 Cross-Survey Multiple Imputation (CSMI)	Pool donor and target data. Impute legal status using multiple imputation (multivariate imputation by chained equations).
5 Logical Cross-Survey Multiple Imputation (Logical-CSMI)	Pool donor and target data. Predicted probability of being unauthorized is coded as 0 for those who are “probably legal,” and this variable is included in the imputation model (multivariate imputation by chained equations).