The PSMATCH Procedure

PSWEIGHT Statement

  • PSWEIGHT <options>;

The PSWEIGHT statement computes weights for the observations on the basis of propensity scores. The PSWEIGHT statement does not apply when you specify either a MATCH statement or a STRATA statement.

Table 6 summarizes the options in the PSWEIGHT statement.

Table 6: PSWEIGHT Statement Options

Option Description
NLARGESTWGT= Displays observations that have the largest weights
WEIGHT= Specifies the type of weight for observations


NLARGESTWGT=n

displays a table of the observations that have the most extreme weights, where n less-than-or-equal-to 50. This option displays observation numbers and weights for the n observations that have the largest weights in the treated and control groups. If you specify an ID statement, the corresponding values of the ID variables are also displayed and serve to identify the observations. By default, n=0 and the table is not displayed.

WEIGHT=ATEWGT | ATTWGT

specifies the type of weight for observations.

ATEWGT <(STABILIZE=YES | NO)>

uses inverse probability of treatment weighting (IPTW). These weights are appropriate for estimation of the ATE. For more information about IPTW-ATE weighting, see the section Inverse Probability of Treatment Weighting.

YES

requests a stabilized inverse probability of treatment weighting.

NO

requests an inverse probability of treatment weighting.

By default, STABILIZE=NO.

ATTWGT

uses ATT weighting (also called weighting by odds) to weight the control group up to the treatment group. These weights are appropriate for estimation of the ATT. For more information about ATT weighting, see the section ATT Weighting.

By default, WEIGHT=ATTWGT.

Last updated: December 09, 2022