I am trying to conduct a logistic regression after propensity score matching. I tried both gmatch psmatch2 and am wondering if I should expect similar result from both? For example :
webuse cattaneo2, clear
rename order orderb
* Using GMATCH
logistic mbsmoke i.mrace i.foreign mage medu fage fedu prenatal orderb i.frace i.mmarried i.fbaby
predict ps
gmatch mbsmoke ps, maxc(1) set(set1) diff(diff1)
logistic deadkids i.mbsmoke i.mrace i.foreign mage medu fage fedu prenatal orderb i.frace i.mmarried i.fbaby if set1 < .
**
* uisng PAMATCH2
psmatch2 mbsmoke i.mrace i.foreign mage medu fage fedu prenatal orderb i.frace i.mmarried i.fbaby, outcome(deadkids) neighbor(1)
logistic deadkids i.mbsmoke i.mrace i.foreign mage medu fage fedu prenatal orderb i.frace i.mmarried i.fbaby [fweight=_weight]
Should I expect similar result for the logistic regression after running the above commands?
Actually in this case the results were close.
************************************************** ************************************************** ************************************************** *****
However, I am doing a similar analysis using another dataset. I run the commands as below:
use mydata, clear
USING GMATCH
logistic disability_child i.healthCond_cat2 i.affectLife i.symptom i.ageCat3 i.workingHourPerWeek_cat4 i.singleFather_n
predict ps
gmatch disability_child ps, maxc(1) set(set1) diff(diff1)
logistic k6cat2v2 i.disability_child i.healthCond_cat2 i.affectLife i.symptom i.ageCat3 i.workingHourPerWeek_cat4 i.singleFather_n if set1 < .
**
USING PSMATCH2
psmatch2 disability_child i.healthCond_cat2 i.affectLife i.symptom i.ageCat3 i.workingHourPerWeek_cat4 i.singleFather_n, outcome(k6cat2v2) neighbor(1)
logistic k6cat2v2 i.disability_child i.healthCond_cat2 i.affectLife i.symptom i.ageCat3 i.workingHourPerWeek_cat4 i.singleFather_n [fweight=_weight]
However, the results are far from similar. Some values are above or below 0, just the opposite. Why is that so?
I want to use psmatch2. Also, I want to produce descriptive table of the matched cases and control after psmatch2. How can I do so?
Related Posts with gmatch and psmatch2
Record linkage with quantitative variablesI have two datasets describing the same 80 schools. Some variables are similar -- for example, both…
exploratory factor analysis w/ maximum likelihood estimation, model comparisonDear Statalisters, I am trying to conduct an EFA on 17 binary variables. I have successfully used b…
Another way of measuring period effect on investment risk tolerance?I'm doing a research about the relationship between investment risk tolerance and demographic factor…
Dummy Variable - Add time variableDear Statalister, I would like to ask how can I create a code for the following variable (See attac…
Flag repetitionsSuppose I have a dataset containing two variables X and Y. Without resorting the data, I would like …
Subscribe to:
Post Comments (Atom)
0 Response to gmatch and psmatch2
Post a Comment