Methods for group comparisons using predicted probabilities and marginal effects on probabilities are developed for regression models for binary outcomes. Unlike approaches based on the comparison of regression coefficients across groups, the methods we propose are unaffected by the scalar identification of the coefficients and are expressed in the natural metric of the outcome probability. While we develop our approach using binary logit with two groups, we consider how our interpretive framework can be used with a broad class of regression models and can be extended to any number of groups.
AgrestiA.2013. Categorical Data Analysis. 3rd ed. New York: Wiley.
2.
AiC.NortonE. C.. 2003. “Interaction Terms in Logit and Probit Models.” Economics Letters80:123–29.
3.
AllisonP. D.1999. “Comparing Logit and Probit Coefficients across Groups.” Sociological Methods & Research28:186–208.
4.
AmemiyaT.1981. “Qualitative Response Models: A Survey.” Journal of Economic Literature19:1483–536.
5.
BenderR.KussO.. 2010. “Methods to Calculate Relative Risks, Risk Differences, and Numbers Needed to Treat from Logistic Regression.” Journal of Clinical Epidemiology63:7.
6.
BishopY.FienbergS.HollandP.. 1975. Discrete Multivariate Analysis: Theory and Practice. Cambridge, MA: MIT Press.
7.
BreenR.HolmA.KarlsonK. B.. 2014. “Correlations and Nonlinear Probability Models.” Sociological Methods & Research43:571–605.
8.
BreenR.KarlsonK. B.. 2013. Counterfactual Causal Analysis and Nonlinear Probability Models. Dordrecht, the Netherlands: Springer.
9.
BuisM. L.2010. “Stata tip 87: Interpretation of Interactions in Non-linear Models.” The Stata Journal10:305–8.
10.
BurbidgeJ. B.MageeL.RobbA. L.. 1988. “Alternative Transformations to Handle Extreme Values of the Dependent Variable.” Journal of the American Statistical Association83:123–27.
11.
ChowG.1960. “Tests of Equality between Sets of Coefficients in Two Linear Regressions.” Econometrica28:591–605.
12.
GreenlandS.1987. “Interpretation and Choice of Effect Measures in Epidemiologic Analyses.” American Journal of Epidemiology125:761–68.
13.
Health and Retirement Study. 2006. Public Use Dataset. Ann Arbor, MI.: Produced and Distributed by the University of Michigan with Funding from the National Institute on Aging (grant number NIA U01AG009740).
14.
HummerR. A.BenjaminsM. R.RogersR. G.. 2004. “Racial and Ethnic Disparities in Health and Mortality among the U.S. Elderly Population.” Pp 53–94 in Critical Perspectives on Racial and Ethnic Differences in Health in Late Life, edited by AndersonNorman B.BulataoRodolfo A.CohenBarney. Washington, DC: National Academies Press.
15.
KendlerK. S.GardnerC. O.. 2010. “Interpretation of Interactions: Guide for the Perplexed.” The British Journal of Psychiatry197:170–71.
16.
KuhaJ.MillsC.. 2018. “On Group Comparisons with Logistic Regression Models.” Sociological Methods & Research1–28.
17.
LandermanL. R.MustilloS. A.LandK. C.. 2011. “Modeling Repeated Measures of Dichotomous Data: Testing Whether the Within-person Trajectory of Change Varies across Levels of Between-person Factors.” Social Science Research40:1456–64.
18.
LeeperT. J. 2018. Margins: An R Port of Statas Margins Command. R package version 0.3.23.
19.
LiaoT. F.2002. Statistical Group Comparison. Vol. 29. New York: Wiley.
20.
LongJ. S.1997. Regression Models for Categorical and Limited Dependent Variables, vol. 7 of Advanced Quantitative Techniques in the Social Sciences. Thousand Oaks, CA: Sage.
21.
LongJ. S.2005. Group Comparisons in Nonlinear Models Using Predicted Outcomes. Working paper, Department of Sociology, Indiana University.
22.
LongJ. S.2009. Group Comparisons in Logit and Probit Using Predicted Probabilities. Working paper, Department of Sociology, Indiana University.
23.
LongJ. S.FreeseJ.. 2006. Regression Models for Categorical Dependent Variables Using Stata. 2nd ed. College Station, TX: Stata Press.
24.
LongJ. S.FreeseJ.. 2014. Regression Models for Categorical Dependent Variables Using Stata. 3rd ed. College Station, TX: Stata Press.
25.
MaddalaG.1983. Limited-dependent and Qualitative Variables in Econometrics. Cambridge, MA: Cambridge University Press.
26.
MarkidesK. S.RudkinL.AngelR. J.EspinoD. V.. 1997. “Health status of Hispanic elderly.” Pp 285–300. in Racial and Ethnic Differences in the Health of Older Americans, edited by MartinL. G.SoldoB. J., Washington DC: National Academy Press.
27.
McKelveyR. D.ZavoinaW.. 1975. “A Statistical Model for the Analysis of Ordinal Level Dependent Variables.” Journal of Mathematical Sociology4:103–20.
28.
MoodC.2010. “Logistic Regression: Why We Cannot Do What We Think We Can Do, and What We Can Do about It.” European Sociological Review26: 1–16.
29.
MustilloS.LandermanL. R.LandK. C.. 2012. “Modeling Longitudinal Count Data: Testing for Group Differences in Growth Trajectories Using Average Marginal Effects.” Sociological Methods & Research41:467–87.
30.
NortonE. C.MillerM. M.KleinmanL. C.. 2013. Computing Adjusted Risk Ratios and Risk Differences in Stata. Stata Journal13:492–509.
31.
NortonE. C.WangH.AiC.. 2004. “Computing Interaction Effects and Standard Errors in Logit and Probit.” The Stata Journal4:154–67.
32.
Þxpert Panel on the Identification, Evaluation, and Treatment of Overweight in Adults. 1998. “Clinical guidelines on the identification, evaluation, and treatment of overweight and obesity in adults: executive summary.” The American Journal of Clinical Nutrition68: 899–917.
33.
RAND. 2014. RAND HRS Data, Version N. Santa Monica, CA: Produced by the RAND Center for the Study of Aging, with funding from the National Institute on Aging and the Social Security Administration.
34.
StataCorp. 2017a. Stata 15 Base Reference Manual. College Station, TX: Stata Press.
35.
StataCorp. 2017b. Stata 15 Survey Data Reference Manual. College Station, TX: Stata Press.
36.
WilliamsR.2009. “Using Heterogeneous Choice Models to Compare Logit and Probit Coefficients across Groups.” Sociological Methods and Research37:531–59.
37.
ZhangQ.WangY.HuangE. S.. 2009. “Changes in Racial/Ethnic Disparities in the Prevalence of Type 2 Diabetes by Obesity Level among US Adults.” Ethnicity & Health14:439–57.
