MatchIt (version 1.0-1)

neyman: Estimating treatment effects: Neyman's method

Description

Calculates basic estimate of treatment effect, using simple difference in means and Neyman's estimate of the variance.

Usage

neyman <- neyman(Y, object, bootstrap=NULL, counter=TRUE)

Arguments

Y

(required). Outcome variable of interest, included in data from matchit object.

object

(required). Stored output from matchit.

bootstrap

Use the bootstrap to estimate standard errors (default=FALSE). Often used when matching done with replacement.

counter

Counter indicating progress of the bootstrap (default=TRUE, which shows the counter).

Value

• Returns estimate of the effect of the treatment, as well as the standard deviation of that estimate.

Details

Calculates overall treatment effect estimate using matched samples generated by matchit. If subclasses were generated in the matching procedure, the estimate is a weighted average over the subclasses, with subclass weights defined by the sub.by option in the matchit call. The standard deviation can also be estimated using the bootstrap procedure, which is often used when matching done with replacement. For more details on the calculations, see the complete documentation (link below). The summary command on a neyman object will provide a test of significance of the outcome as well as sample sizes of the matched treated and control groups, and subclass estimates if applicable.

See Also

Please use help.matchit to access the matchit reference manual. The complete document is available online at http://gking.harvard.edu/matchit.