Cem March 2016

cem@lists.gking.harvard.edu

2 participants
3 discussions

by 서연수

Hello, I had two questions regarding using the CEM package in Stata that I would be very greatful if someone helped me with. 1) In the 2009 Stata journal (cem: Coarsened exact matching in Stata), specifically the part of 4.7 Using cem to improve other matching methods, it says that after CEM, PSM(propensity score matching) can be used to obtain better inferences. I was wondering if this applied to propensity score weighting as well as propensity score weighting has the disadvantage of producing "bad" weights for outliers I believe. (In short, I want to know if I can conduct CEM and then using the data, conduct propensity score weighting-in order to get rid of much bias as possible) Moreover, if this works, would I add the same covariates used in CEM and propensity score weighting in the final analysis model? 2) Also, when using CEM, the weights are incorporated in following analysis using the "iweights" option in Stata, However, in propensity score weigthing, "pweights" seem to be the norm. Is there a reason why the two methods use different weights? The results of iweights and pweights differ for both methods and I was confused. I am not familiar with weights, so I might be asking a silly question but I was curious. I know it is a lot of questions, but would really appreciate it if someone could help me out. Thank you in advance.

8 years

Re: [cem] Stata Output from CEM - matched observations with missing strata

by Moulton, Stephanie

Did you ever find a response to the question about STATA Output from CEM- matched observations with missing strata? We are finding the same problem with our data. None of the matching variables have missing data, yet many of the observations are being assigned to a "missing" stratum (complete with CEM weights for the missing stratum). Any insights? Thanks! Stephanie Moulton Associate Professor and Director of Doctoral Studies John Glenn School of Public Affairs The Ohio State University http://glenn.osu.edu/faculty/glenn-faculty/moulton/

8 years

cem for spss: k2k not working in 22/ installation problems in 21

by Ditzhuijzen, J.M. van (Jenneke)

Dear Matt and/or others, For weeks I have been struggling to install and use CEM for SPSS. I do not know R, and am not familiar with programming, so I am very sorry if I ask any stupid questions.. I have been able to use CEM before in SPSS 20, and it worked out very well, but am now not able to reproduce this in SPSS 21 or 22. I really want to finish these analyses, so I would be very grateful if someone could advice me. So I have two questions, solving one of them would be enough for me :) I tried in SPSS 22: I managed to install CEM. But the k2k option is not working. Something seems to go wrong with the RandomForest option. And I cannot see the error messages, so I really don't know what to do. I have pasted the output below, perhaps this looks familiar to someone. I think the problem is SPSS related, not CEM, but am not sure. I have downloaded the version of R associated with RandomForest; so I don't think this could be the problem (but anything's possible). I tried in SPSS 21 on my old laptop. I cannot fully install CEM, I get an error message while installing the SPSS Essentials for Python 21 plugin, it says it needs SPSS version 21.0., but it can only find SPSS 21.0.??? See attached photo. If someone could be so kind to advice me, I would be very happy :) Thanks a lot!! Best, Jenneke (NL) DATASET NAME DataSet1 WINDOW=FRONT. CEM TREATMENT=xcohort VARIABLES=xagecat2 xnopartner xnon_west xchildren xcity xnojob xanymisbr /OPTIONS K2K NOIMBAL. Loading required package: nlme Loading required package: lattice Loading required package: randomForest randomForest 4.6-10 Type rfNews() to see new features/changes/bug fixes. Loading required package: combinat Attaching package: 'combinat' The following object(s) are masked from 'package:utils': combn How to use CEM? Type vignette("cem") Warning messages: 1: package 'cem' was built under R version 2.15.3 2: package 'randomForest' was built under R version 3.2.0 Matching Summary G1 G2 All 1888 325 Matched 1566 311 Unmatched 322 14 The data contain missing values. CEM will match on them; see the manual for other options. There were 50 or more warnings (use warnings() to see the first 50) warnings(). Error # 1. Command name: warnings The first word in the line is not recognized as an SPSS Statistics command. Execution of this command stops.

8 years, 1 month

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

Cem March 2016