Statistical Analysis Stack Exchange
Statistical Analysis Stack Exchange

Questions tagged [marginal-model]

A marginal model estimates population averages, in contrast to a subject-specific model that estimates an expected value conditional on a subject's attributes. In a linear model context, these are the same, but they differ for nonlinear models. The GEE is the most salient marginal model.

A marginal model estimates population averages, in contrast to a subject-specific model that estimates an expected value conditional on a subject's attributes. In a linear model context, these are the same, but they differ for nonlinear models. The GEE is the most salient marginal model.

Observations coming from the same unit are not (statistically) independent. This poses particular challenges to the statistical procedures used for the analysis of such data. A marginal model (e.g., the Generalized Estimating Equations approach) is a method to account for such nonindependent, autocorrelated data. Contrary to a (subject specific) GLMM, a marginal model yields parameters with a population-averaged interpretation.

Reference: Verbeke et al. (2010). Random Effects Models for Longitudinal Data. Springer.

26 questions
15
votes
1 answer

The difference between average and marginal treatment effect

I have been reading some papers, and I am unclear about the specific definitions of Average Treatment Effect (ATE), and Marginal Treatment Effect (MTE). Are they the same? According to Austin... A conditional effect is the average effect, at the…
RayVelcoro
  • 1,039
  • 1
  • 10
  • 19
13
votes
1 answer

Marginal model versus random-effects model – how to choose between them? An advice for a layman

In searching for any info about marginal model and random-effects model, and how to choose between them, I have found some info but it was more-or-less mathematical abstract explanation (like for example here:…
benjamin jarcuska
  • 369
  • 1
  • 4
  • 16
10
votes
1 answer

Difference between marginal and conditional models

A marginal model accounts for the correlation within each cluster. A conditional model also takes into account the correlation within each cluster. My questions are: Does a marginal model models main effects across a population whereas a…
jamie
  • 131
  • 1
  • 1
  • 3
4
votes
2 answers

how to replicate the stata "margins, atmeans" command in R with the margins library

i'm trying to replicate the output of margins female, atmeans in R shown here: https://stats.idre.ucla.edu/stata/dae/using-margins-for-predicted-probabilities/ i can re-create the same setup shown on the ucla page…
Anthony Damico
  • 272
  • 2
  • 17
3
votes
1 answer

Subject specific vs population average predictions

I am in doubt whether in my thesis I should report on the subject specific predictions of the probability to respond with an 'I don't know' answer, or the population average. Consider for example the following graphs (I have not gotten them smaller…
Marloes
  • 740
  • 6
  • 21
3
votes
2 answers

emmeans: command emmeans() or contrast(), whats the difference?

I hope somebody is available to help a desperate rookie.. I fitted a glmer with a Poisson distribution and log link, including main effects and several interactions, an offset variable and a random effect. Something like this: model1 <- glmer(count…
Ronja
  • 41
  • 4
3
votes
0 answers

How to control for repeated measurements without controlling for between-cluster variability?

I want to explore the effect of the average prenatal maternal stress (cortisol level, continuous measure) on offspring growth during a linear growth period (monthly body size measure, N = 17 infants, 16.7+-1.3 body size measures per infant). So my…
Andreas
  • 31
  • 1
2
votes
0 answers

When to choose mixed effect model over marginal model?

I am reading Applied Longitudinal Data Analysis by Fitzmaurice. However, I am confused as when to use mixed effect model. The book says in general when we want to study impact on individuals of the study, we apply mixed and when we want to study…
user45765
  • 765
  • 3
  • 10
2
votes
1 answer

Adding baseline covariates in stabilized weights change covariate balancing

I am computing weights using inverse probability of treatment weighting for marginal structural models (Robins et al. 2000). With both time-varying and time-invariant (baseline) covariates, some papers use baseline covariates in the computation of…
Maël
  • 187
  • 6
2
votes
0 answers

Inverse probability of treatment weights and linear mixed effects models

I am encountering a problem when using inverse probability of treatment weights with linear mixed-effects models for a difference-in-differences analysis. I have longitudinal data on participants. I weighted each treatment condition to be equal on…
ATB
  • 21
  • 2
2
votes
1 answer

Getting a 0 correlation for Poisson marginal model in geepack in R

I'm trying to replicate Table 13.8 from Fitzmaurice, Laird, & Ware (2011) using R for teaching purposes. This is a GEE count model of the number of bacteria on 30 patients at two waves. In their SAS-estimated results, they get a correlation between…
Jamie Monogan
  • 23
  • 4
1
vote
1 answer

Multivariate Regression with Two Different Types of Response

Problem Setting: I have an interesting question related with longitudinal study and multivariate regression. I found that in lots of biomedical studies, multiple discrete and continuous endpoints are very common. For example, continuous fetal weight…
KPMGGMC
  • 21
  • 3
1
vote
1 answer

Do Marginal Models by nature not have enough degrees of freedom (and therefore cannot fit)?

In the following text from Agresti's Foundations of Linear and Generalized Linear Models, I just don't get how equation 9.2 makes any sense. We are making a separate linear relationship between each student and their jth exam and the response. This…
Aegis
  • 133
  • 4
1
vote
1 answer

Interpreting log linear margins with endogenous treatment effects

I'm having trouble in understanding the predictive margins after a log linear regression with endogenous treatment effects. Using stata (with weighted survey design) I ran the following, where logwage is the log of wage. The log was taken because…
iPlexipen
  • 211
  • 2
  • 7
1
vote
0 answers

GEEs cannot analyze demographic variables but GLMMs can?

I thought I was coming to understand the difference between (binary) GLMMs and Marginal Models using GEE...until I encountered the following passage in Hosmer et al (2013: 328): The clear weakness of the population average model is that it cannot…
blokeman
  • 136
  • 1
  • 8
1
2