Interpretation of Odds Ratios in xtlogit for Panel Data

[Nermine Medhioub]

Hello,

I am running a logistic regression for panel data using xtlogit in Stata. My dependent variable is binary, where 1 indicates the occurrence of the event and 0 indicates its non-occurrence.

I have a question regarding the interpretation of the odds ratios produced by the model. Specifically, I am trying to understand whether these odds ratios reflect the odds of the event occurring (i.e., P(event) / P(non-event)), or the odds of the event not occurring (i.e., P(non-event) / P(event)).

To clarify, when I look at the odds ratios:

• Is an odds ratio > 1 indicating that the predictor increases the probability of the event occurring (1 in the numerator and 0 in the denominator)?
• Or, is it the opposite, as I have noticed in SPSS, where the event (1) is in the denominator?

Thank you for your help!

[George Ford]

https://stats.oarc.ucla.edu/spss/library/spss-libraryunderstanding-odds-ratios-in-binary-logistic-regression/

sysuse auto, clear
reg foreign mpg
margins, at(mpg = (20 21))
logit foreign mpg , or

[Nermine Medhioub]

@George Ford the interpretation of log odd ratio in stata is it the same as in spss?

[Nermine Medhioub]

@Joao Santos Silva

[George Ford]

An odds ratio is an odds ratio. It has a definition. All stat packages will interpret them the same (#2 shows both Stata and SPSS are the same), though I suspect there may be options to recast them if requested.

[Nermine Medhioub]

thank you @George Ford