DESIGN =JACKKNIFE is often used in non-survey applications involving clustered data. A common Jackknife method for sample surveys is to delete one PSU (or cluster for correlated data in non-survey applications). The sampling weights for the remaining PSUs in the same stratum are adjusted to account for the deleted PSU. We refer to this as the Delete-1 Jackknife method, and it is the default Jackknife design in SUDAAN.
For this design the WEIGHT and NEST statements are required. The default NEST statement specification for the JACKKNIFE design is:
NEST STRATA PSU;
where STRATA defines the stratification variable, and PSU defines the primary cluster variable. To define the entire population as a single stratum, use the SUDAAN keyword _ONE_ in place of STRATA.
If there are many PSUs, it is possible to construct a large number of Jackknife replicates. SUDAAN make this construction possible through two sample design statements, JACKWGTS and JACKMULT. We refer to this option as the Replicate Weight Jackknife method. For this design the WEIGHT statement is optional, the JACKWGTS statement is required, the JACKMULT statement is optional, and the NEST statement is used only in the modeling procedures to indicate the cluster variable, when R=EXCHANGEABLE is specified.
For replicate weight Jackknife designs,
You may use the REPDATA option on the PROC statement to specify a separate file of replicate weights. In this case, you must also include an IDVAR statement to specify the synchronizing variable(s) on the REPDATA file. You may also include a REP_REC option on the PROC statement to specify the number of records on the REPDATA file.