Threshold of Covariate Balance Check on Nearest Match

[Yasutake Homma]

Hello Statalist,

I would like to ask you about the threshold of Covariate Balance Check on Nearest Match.

After:
teffects nnmatch

I used

tebalance sum

Then, I will find Standardized differences of raw and matched.
Some website told me the threshold for good-match should be +-0.25.

Is this based on an academic papers so that I can use the threshold for my academic paper which would be submitted to international journal?

If you could give me any information, it would be appreciated.
Thank you.

Regards;
HOMMA

[Rich Goldstein]

There are a number of criteria by various people in the lit; that specific one, and a "companion" for the variance ratios, can be found in Stuart, EA and Rubin, DB (2011), "Best practices in quasi-experimental designs: matching methods for causal inference", in Best practices in quantitative methods, ed. by J Osborne, Sage Publications (p. 19); some others are discussed in, e.g., Bai, H and Clark, MH (2019), Propensity score methods and applications, Sage Quantitative applications in the social sciences 178

[Yasutake Homma]

Thank you very much and sorry for my late reply.