The parameters of logit models are typically difficult to interpret, and the applied literature is replete with interpretive and computational mistakes. In this article, I review a menu of options to interpret the results of logistic regressions correctly and effectively using Stata. I consider marginal effects, partial effects, (contrasts of) predictive margins, elasticities, and odds and risk ratios. I also show that interaction terms are typically easier to interpret in practice than implied by the recent literature on this topic.

Uberti, L. (2022). Interpreting logit models. THE STATA JOURNAL, 22(1), 60-76 [10.1177/1536867x221083855].

Interpreting logit models

Uberti, LJ
2022

Abstract

The parameters of logit models are typically difficult to interpret, and the applied literature is replete with interpretive and computational mistakes. In this article, I review a menu of options to interpret the results of logistic regressions correctly and effectively using Stata. I consider marginal effects, partial effects, (contrasts of) predictive margins, elasticities, and odds and risk ratios. I also show that interaction terms are typically easier to interpret in practice than implied by the recent literature on this topic.
Articolo in rivista - Articolo scientifico
st0663; logit; binary outcome models; nonlinear models; interpretation; interaction terms
English
2022
22
1
60
76
reserved
Uberti, L. (2022). Interpreting logit models. THE STATA JOURNAL, 22(1), 60-76 [10.1177/1536867x221083855].
File in questo prodotto:
File Dimensione Formato  
Uberti-2022-The Stata J-VoR.pdf

Solo gestori archivio

Descrizione: Article
Tipologia di allegato: Publisher’s Version (Version of Record, VoR)
Licenza: Tutti i diritti riservati
Dimensione 406.73 kB
Formato Adobe PDF
406.73 kB Adobe PDF   Visualizza/Apri   Richiedi una copia

I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: https://hdl.handle.net/10281/446608
Citazioni
  • Scopus 7
  • ???jsp.display-item.citation.isi??? 7
Social impact