Re: [R] P-value from Matching

2015-02-16 Thread Rob Wood
Hi Peter, That was my first port of call before I posted this thread. Unfortunately, it does not seem to explicitly state which test is used or how the p-value is calculated. Thanks, Rob. -- View this message in context: http://r.789695.n4.nabble.com/P-value-from-Matching-tp4703309p4703345.h

[R] P-value from Matching

2015-02-15 Thread Rob Wood
Hi all, When using the match command from the matching package, the output reports the treatment effect, standard error, t-statistic and a p-value. Which test is used to generate this p-value, or how us it generated? Thanks, Rob. -- View this message in context: http://r.789695.n4.nabble.com

[R] GenMatch providing different results

2014-12-16 Thread Rob Wood
Hi there, I'm performing some propensity score matching in R, where I am using GenMatch() to generate a weight matrix for the covariates, before plugging that into Match(). However, I keep on getting different results each time I run the code. I'm using int.seed and unif.seed in GenMatch, but to n