Matchit January 2016

matchit@lists.gking.harvard.edu

4 participants
5 discussions

by Yi Xe Thng

Sorry for the multiple posts. I am reposting the message as the formatting didn't appear right. Given the following notation, do I calculate the ATE weights as below? Notation: n = overall sample size, ni = sample size within stratum i, nt = no. of treatment subjects in overall sample, nc = no. of comparison subjects in overall sample, nti = no. of treatment subjects within stratum i, and nci = no. of comparison subjects within stratum i. Is the ATE weights for the treatment and subject in stratum i respectively: ATEweight_ti = ( ni / nti ) * (nt / n ) ATE weight_ci = ( ni / nci ) * (nc / n )? By extension, would this also be the way to calculate the ATE weights after subclassification? Many thanks, Yi Xe

8 years, 3 months

Data warning message for full matching

by Yi Xe Thng

Hi, I ran a command for full matching, but keep getting a warning message. Is there something missing in the way I wrote the command, or do I need to set up the data in a special way for full matching? Also, what are the implications of ignoring the warning? *Command:* m.out <- matchit (treat ~ X1 + X2 + X3 + X4 + X5 + ...+Xn , data = df4w1, method = "full", distance = "logit") summary(m.out, standardize = TRUE) *Warning message:* In fullmatch(d, ...) : Without 'data' argument the order of the match is not guaranteed to be the same as your original data. Many thanks, Yi Xe

8 years, 3 months

ATE for Full Matching

by Yi Xe Thng

Hi, I am trying to calculate the ATE weights after conducting full matching. There was a reference to the following documentation for the calculation of ATT weights, but the access was denied: https://lists.gking.harvard.edu/pipermail/matchit/2011-September/000312.html Given the following notation: n = overall sample size, n*i *= sample size within stratum *i*, n*t* = no. of treatment subjects in overall sample, n*c* = no. of comparison subjects in overall sample, n*ti* = no. of treatment subjects within stratum *i*, and n*ci* = no. of comparison subjects within stratum *i*, Is the ATE weights for the treatment and subject in stratum *i* respectively: ATEweight*ti* = ( n*i* / n*ti* ) * ( n*t* / n ) ATEweight*ci* = ( n*i* / n*ci* ) * ( n*c* / n ) ? By extension, would this also be the way to calculate the ATE weights after subclassification? Many thanks, Yi Xe

8 years, 3 months

Output piared data set naming

by William Dudley

I am using MatchIT through an SPSS shell called* Propensity Score Matching for SPSS version 3.0.4 by Felix Thoemmes*. I like the SPSS shell because it provides a seamless process for me. I have two questions: First, I am matching schools in two states and need to run the matching twice once for each state. I am wondering if there is a way to run matchit with a strata for states so that I can get balance statistics for each state separately as well as balance with all school combined across states. Second, The output provides the paired data set and names it with a random number. This makes it difficult to save the paired data using syntax (it is named differently on each run). I understand that the random number reduces the likelihood that files will be over written however, I am wondering if it is possible to set the random seed so that the file name remains constant across runs. Thanks Bill -- William N. Dudley, PhD Professor - Public Health Education The School of Health and Human Sciences The University of North Carolina at Greensboro 437-L Coleman Building Greensboro, NC 27402-6170 Visit my Web Site <http://www.uncg.edu/phe/faculty/dudley.html> VOICE 336.256 2475

8 years, 3 months

Error in mahvars[pool, , drop = F] : subscript out of bounds

by Rooney, Patrick

I have the same error using the mahvars option with method="nearest" as noted by Yuqin Wei weiyuqin at gmail.com Tue Sep 9 20:13:30 EDT 2014 [matchit] Bug Using 'mahvars' Option My data: > str(PrePost5) 'data.frame': 59425 obs. of 16 variables: $ wallet : int 281230572 301242637 246610352 109763234 283176785... $ debut : int 1 1 1 1 1 1 1 1 1 1 ... $ age : num 37.2 40.5 54.3 64.8 16 ... $ sales : num 476.3 46.5 576.8 32.9 41.4 ... $ trips : int 8 4 8 1 2 1 2 2 1 9 ... $ FLS_pre_ind : int 0 1 0 0 1 1 1 1 0 1 ... $ Ncom_pre_ind : int 0 0 1 0 0 0 0 0 1 1 ... $ trips_Nov2014 : int 0 0 0 0 1 1 1 1 1 1 ... $ gender_F_ind : int 1 1 1 1 1 1 1 0 0 1 ... $ Fchan_FC_flag : int 0 0 0 0 0 0 0 0 1 1 ... $ Fchan_HL_flag : int 0 0 0 0 0 0 0 0 0 0 ... $ Nfsb.Adoption.Rate: int 0 0 0 0 0 0 0 0 0 0 ... $ Email : int 0 0 0 0 0 0 0 0 0 0 ... $ pre_max_pb_flag : int 0 1 0 0 0 0 0 0 0 1 ... $ Recency_weeks : num 3 3 1 3 1 3 0 2 3 3 ... $ NumYrs : num 4.75 5.66 2.03 3.96 1.89 1.56... The data.frame PrePost5 appears to have row.names, correct? (previous post from Wei indicated that might be a problem): row.names(PrePost5) [1] "1" "2" "3" "4" "5" "6" "7" "8" "9" "10" "11" "12" Call to MatchIt and error message: m.out.nn <- matchit(debut ~ age + sales + trips +FLS_pre_ind+Ncom_pre_ind+trips_Nov2014+ gender_F_ind+Fchan_FC_flag+Fchan_HL_flag+Email+pre_max_pb_flag+Recency_weeks+NumYrs, data = PrePost5,method = "nearest", exact=c("FLS_pre_ind","Ncom_pre_ind","gender_F_ind","Fchan_FC_flag","Fchan_HL_flag", "Email","pre_max_pb_flag"), mahvars=c("age","sales","trips","trips_Nov2014","Recency_weeks","NumYrs")) Error in mahvars[pool, , drop = F] : subscript out of bounds In addition: Warning messages: 1: glm.fit: fitted probabilities numerically 0 or 1 occurred 2: No caliper size specified for Mahalanobis matching. Caliper=.25 used. Sorry if I didn't see the reply to the Sep 2013 reporting of same error. Any help would be greatly appreciated!

8 years, 3 months

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

2006

2005

2004

Matchit January 2016