The PSMATCH Procedure

OUTPUT Statement

OUTPUT OUT <(OBS=obs-value )>=SAS-data-set <keyword=name <keyword=name …>> ;

The OUTPUT statement specifies the output data set and variables. You must specify the following option:

OUT < (OBS=obs-value )>=SAS-data-set

names the output data set. The data set also includes the results of matching if you provide the MATCH statement. You can specify one of the following values for obs-value:

ALL: the output data set contains all observations.
REGION: the output data set contains only observations in the specified support region.
MATCH: the output data set contains only the matched treated units and control units. This option applies only if you specify the MATCH statement.

By default, OBS=ALL.

You can also specify the one or more of the following keywords to create and name the output variables:

ATEWGT=name

creates and names the weight variable that provides inverse probability of treatment weighting. This weighting is appropriate for estimating the ATE.

ATTWGT=name

creates and names the weight variable for ATT weighting. This weighting is appropriate for estimating the ATT. If ATTWGT= is not specified and neither the MATCH nor the STRATA statement is specified, then this variable is automatically created with name=_ATTWGT_.

LPS=name

creates and names the variable that provides the logit of propensity score.

MATCHID=name | (names )

creates and names the variable that provides identification numbers for the matched treated and control units. This suboption applies only if you also specify the MATCH statement. If METHOD=REPLACE(K=k) is specified with k > 1, then you can use MATCHID=(names) to name the k matching groups for each treated unit.

MATCHWGT=name

creates and names the weight variable for the matching. This suboption applies only if you also specify the MATCH statement. In each matched set, each treated unit has a weight of 1 and each control unit has a weight that equals the number of treated units divided by the number of control units in the matched set. With this weighting, the total weight of control units is the same as the total number of treated units in each matched set, and the total weight of matched control units is the same as the total number of treated units. This weighting is appropriate for estimating the ATT.

If MATCHWGT= is not specified but the MATCH statement is specified, then this variable is automatically created with name=_MATCHWGT_.

PS=name

creates and names the variable that provides the propensity score.

If PS= is not specified and the PS= option in the PSDATA statement is also not specified, then this variable is automatically created with name=_PS_.

STRATA=name

creates and names the variable that numbers the strata. The suboption applies only if the STRATA statement is specified.

If STRATA= is not specified but the STRATA statement is specified, then this variable is automatically created with name=_STRATA_.