Best Practices for Estimating, Interpreting, and Presenting Nonlinear Interaction Effects

2019/01/01 by Trenton D. Mize · 1 voice · 41 citations
Mathematics · Social Sciences · #Advanced Causal Inference Techniques #Advanced Statistical Methods and Models #Income, Poverty, and Inequality

paper · doi:10.15195/v6.a4

openalex publication_date 2019/01/01 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/25

Abstract

Many effects of interest to sociologists are nonlinear. Additionally, many effects of interest are interaction effects—that is, the effect of one independent variable is contingent on the level of another independent variable. The proper way to estimate, interpret, and present these two types of effects individually are well known. However, many analyses that combine these two—that is, tests of interaction when the effects of interest are nonlinear—are not properly interpreted or tested. The consequences of approaching nonlinear interaction effects the way one would approach a linear interaction effect are severe and can often result in incorrect conclusions. I cover both nonlinear effects in the context of linear regression, and—most thoroughly—nonlinear effects in models for categorical outcomes (focusing on binary logit/probit). My goal in this article is to synthesize an evolving methodological literature and to provide straightforward advice and techniques to estimate,interpret, and present nonlinear interaction effects.

Citations

• Mechanisms Linking Social Ties and Support to Physical and Mental Health
• Making the Most of Statistical Analyses: Improving Interpretation and Presentation
• Interpreting and Understanding Logits, Probits, and Other Nonlinear Probability Models

Cited by

• Determinants of Frequent Jail Admissions: Evidence From Two Counties
• Resident funding and care home quality: a retrospective observational analysis of the impact of the two-tier care system in England
• Religiosität und Delinquenz unter christlichen und muslimischen Jugendlichen
• The code of cohesion: adolescent network centrality, offending, and the downside of school cohesion
• From Victim to Offender: The Role of Dark Personality Traits in School Violence
• Marginal Odds Ratios: What They Are, How to Compute Them, and Why Sociologists Might Want to Use Them
• The Transformation of Polygyny in Sub‐Saharan Africa
• Political Identity and Confidence in Science and Religion in the United States
• Short-term flooding in non-rice croplands provides stopover habitats for migrating waterbirds
• Incivility in Comparison: How Context, Content, and Personal Characteristics Predict Exposure to Uncivil Content
• An imperfect match? gender and racial discrimination in hiring and applicant-job requirement matching as an unequal burden of proof
• Compensatory and Multiplicative Advantages: Social Origin, School Performance, and Stratified Higher Education Enrolment in Finland
• Transmorphic organizations: racial segregation and discrimination against LGBTQ students at Christian colleges and universities
• Crime propensity and lifestyle risk: The interplay of personal morality and self-control ability in determining the significance of criminogenic exposure
• Roots, Belonging, and Digital Infrastructure: Determinants of Crisis Giving Among Somalis in Eastleigh, Nairobi
• For what benefit? State Right to Work laws and employer-provided retirement and health insurance benefits
• Environment and the racialization of space in US cities
• The Impact of Mail, Web, and Mixed-Mode Data Collection on Participation in Establishment Surveys
• Duration of Marriage Exacerbates the Likelihood of Spousal Violence among Involuntarily Childless Women in Nigeria: Evidence from a National Survey
• Anxiety coping motives moderate links between discrimination and alcohol use in immigrant‐origin and non‐immigrant LGBQ + college student
• Intergenerational occupational mobility and health in the United States
• Cómo interpretar regresiones logísticas binarias con Odds Ratio Marginales y Promedios de Efectos Marginales: una guía aplicada con Stata
• Party Youth Wings as Forces of Renovation: A Study of Young Women Members’ Efficacy and Ambition
• Temporal stability and volatility in population-level high-risk illicit polysubstance use by gender and marital status in the United States, 2002–2019
• Do updates of group-level and individual-level information of ethnic minority applicants mitigate statistical discrimination in hiring?
• Childbearing attitudes amid successive novel infectious disease epidemics: a mixed-method experimental approach
• Fathers Combining Work and Care: Flexible Work Arrangements and Paternal Involvement Across Financial Situations
• Sanctified Bets? Religiosity, Gender, and Sports Gambling in the United States
• Gender and climate change views in context: a cross-national multilevel analysis
• Perceived Social Isolation Among Older Lesbian, Gay, Bisexual, and Heterosexual Adults: The Role of Relationship Status
• Leaving for life: using online crowd-sourced genealogies to estimate the migrant mortality advantage for the United Kingdom and Ireland between the 18 th and 20 th centuries
• Do parenthood worries impede the birth of a second child? Differences according to the parent’s gender and spousal support in Finland
• Patterns of university progression and social inequalities: delving into complex trajectories in higher education
• Violence and Competition: The Effect of Mass Shootings and Domestic Terrorism on Organizational Risk-Taking in Response to Performance Shortfalls
• Who prefers smartphone surveys? Revisiting predictors of survey mode preference
• Assessing Racial Heterogeneity in “Housing First” Supports’ Effectiveness Among Older Adults Experiencing Homelessness: Evidence From Los Angeles County
• Benefits of increasing the value of respondent incentives during the course of a longitudinal mixed-mode survey
• Web and/or MD?: Empirically testing the relationships between internet use and visits to healthcare professionals
• Evaluating the causal mechanisms from procedural justice to police legitimacy and compliance in the United States: An in‐depth mixed‐methods approach
• Gone too long or back too soon? Perceptions of paid parental leave‐taking and variations by gender and family structure
• Radical right support and the deepened rural-urban and educational divides among younger generations

Discussions

• Long, J. S., & Mustillo, S. A. (2018). Using Predictions and Marginal Effects to Compare Groups in Regression Models for Binary Outcomes. Sociological Methods & Research. doi.org/10.1177/0049... 3/ [bsky, 5 points, 1 comments]

Related

• Interpreting Interaction Effects in Generalized Linear Models of Nonlinear Probabilities and Counts
• Anything Goes: Statistical Interactions Without Substantive Theory
• To interact or not to interact: The pros and cons of including interactions in linear regression models
• Probing Interactions in Fixed and Multilevel Regression: Inferential and Graphical Techniques
• Total, Direct, and Indirect Effects in Logit and Probit Models