How to illustrate the result of teffects psmatch for binary outcome?

[Eva Di]

Hi,

I am currently trying to do propensity score matching. The code works well but I have some problem about how to explain the result.

The example is:
use "http://www.stata-press.com/data/r13/cattaneo2.dta", clear
teffects psmatch (lbweight) (mbsmoke mmarried alcohol mage medu,logit)

The result is:

My understanding of the ATE(.431315) here is the difference of odd ratios between smoker and nonsmoker. Is that correct?

If not, how to explain it correctly?

Thanks for your time!

[Joerg Luedicke (StataCorp)]

My understanding of the ATE(.431315) here is the difference of odd ratios between smoker and nonsmoker. Is that correct?

No. The average treatment effect from the propensity score matching estimator is the average of the differences between observed and potential outcomes. If the outcome variable is binary, then this average can be interpreted as a difference in probability. In this example, the result of ATE=0.043 indicates that smokers have a higher probability of giving birth to a low birth weight baby than non-smokers, and the average difference in probability is 0.043. Expressing this result on a percentage scale, we could say that the chance of giving birth to a low birth weight baby is higher by 4.3 percentage points for smokers, compared to non-smokers (notice that this is a difference in percentage points, not a relative difference in percent).

Joerg

[Eva Di]

Thank you so much, Joerg!

[Zhu Liu]

I have a related question. Instead of opening a new thread, I'd like to ask it here so that people can see these basic questions together.

The default output only presents the difference in the outcome variable between the treatment and matched groups. For illustration, I'd like to also report the average outcome for the treatment group and that for the matched group as well. Is there a reporting option that directly report those?

I'd also like to check the characteristics of the treated and the matched and see if there is any significant difference in any of the matching dimensions. Do I need to compute that myself using the output from "tebalance summarize X Y Z, baseline" or is there a command or option doing that directly?

And for both questions, do I need to create the matched pairs and use that to do the two tasks? Is that the only way to do it?

Any help is greatly appreciated.

ZL

[Jae Li]

@Zhu Liu Hi Liu, I also have the same question as you mentioned above, have you found the solutions yet? I will be really appreciated if you can advise. Many thanks in advance!