Fitting Multilevel Models When Predictors and Group Effects Correlate

@article{Bafumi2007FittingMM,
  title={Fitting Multilevel Models When Predictors and Group Effects Correlate},
  author={Joseph Bafumi and Andrew Gelman},
  journal={Econometrics: Econometric \& Statistical Methods - General eJournal},
  year={2007}
}
  • Joseph Bafumi, A. Gelman
  • Published 3 September 2007
  • Psychology
  • Econometrics: Econometric & Statistical Methods - General eJournal
Random effects models (that is, regressions with varying intercepts that are modeled with error) are avoided by some social scientists because of potential issues with bias and uncertainty estimates. Particularly, when one or more predictors correlate with the group or unit effects, a key Gauss-Markov assumption is violated and estimates are compromised. However, this problem can easily be solved by including the average of each individual-level predictors in the group-level regression. We… 
Should I Use Fixed or Random Effects?
Empirical analyses in social science frequently confront quantitative data that are clustered or grouped. To account for group-level variation and improve model fit, researchers will commonly specify
Multilevel Modeling and Ordinary Least Squares Regression: How Comparable Are They?
ABSTRACT Studies analyzing clustered data sets using both multilevel models (MLMs) and ordinary least squares (OLS) regression have generally concluded that resulting point estimates, but not the
Explaining Fixed Effects: Random Effects Modeling of Time-Series Cross-Sectional and Panel Data*
This article challenges Fixed Effects (FE) modeling as the ‘default’ for time-series-cross-sectional and panel data. Understanding different within and between effects is crucial when choosing
Multilevel Modeling Myths
  • Francis L. Huang
  • Psychology
    School psychology quarterly : the official journal of the Division of School Psychology, American Psychological Association
  • 2018
TLDR
Although nesting should be accounted for, researchers should realize that MLM, which is a powerful and flexible technique, is not the only method that can be used to account for the clustering effect.
Latent Factor Regressions for the Social Sciences
TLDR
It is shown that interactive latent factor models provide a powerful modeling alternative that can address a wide range of data types and introduce a class of fast variational inference algorithms that allows for models to be fit quickly and accurately.
Two Multilevel Modeling Techniques for Analyzing Comparative Longitudinal Survey Datasets*
TLDR
Two simple techniques for extracting insights into the political, social and economic correlates of many individual-level outcomes of interest by allowing change over time in y to be a function of changeover time in x and/or of a time-invariant x.
Augmented Weighted Estimators Dealing with Practical Positivity Violation to Causal inferences in a Random Coefficient Model
TLDR
The augmented estimating function can be shown to converge in expectation to zero and therefore yield consistent causal estimates and it performs well in simulated data as well as with real-world data from a teacher preparation evaluation study.
R AND BIAS REDUCTION IN MATCHING HIERARCHICAL DATA: A MONTE CARLO STUDY
Based on a two-level structural equation model, this simulation study examines how omitted variables affect estimation bias in matching hierarchical data. Six simulated cases of omitted variables are
Practical and Effective Approaches to Dealing With Clustered Data
TLDR
Develop new and easy-to-use Stata and R packages that implement alternative uncertainty measures robust to small G, and explain and provide evidence for the advantages of these alternatives, especially cluster-adjusted t-statistics based on Ibragimov and Müller.
A Bayesian Multilevel Modeling Approach to Time-Series Cross-Sectional Data
TLDR
Monte Carlo simulations are employed to benchmark the performance of a Bayesian multilevel model for TSCS data and find that the MLM performs as well or better than other common estimators for such data.
...
...

References

SHOWING 1-10 OF 17 REFERENCES
Modeling Multilevel Data Structures
TLDR
The logic and statistical theory behind multilevel models are introduced, to illustrate how such models can be applied fruitfully in political science, and to call attention to some of the pitfalls in multileVEL analysis.
Bayesian Multilevel Estimation with Poststratification: State-Level Estimates from National Polls
TLDR
The most important usage of this method is envisioned to be not forecasting elections but estimating public opinion on a variety of issues at the state level.
A Bayesian Multilevel Modeling Approach to Time-Series Cross-Sectional Data
TLDR
Monte Carlo simulations are employed to benchmark the performance of a Bayesian multilevel model for TSCS data and find that the MLM performs as well or better than other common estimators for such data.
Random Coefficient Models for Time-Series—Cross-Section Data: Monte Carlo Experiments
This article considers random coefficient models (RCMs) for time-series—cross-section data. These models allow for unit to unit variation in the model parameters. The heart of the article compares
Pooling Disparate Observations
TLDR
Fractional pooling provides a simple and coherent way to incorporate prior beliefs about the theoretical relevance of disparate observations or to explore the implications of prior uncertainty about their relevance, and has a plausible rationale in Bayesian statistical theory.
Poststratification Without Population Level Information on the Poststratifying Variable With Application to Political Polling
We investigate the construction of more precise estimates of a collection of population means using information about a related variable in the context of repeated sample surveys. The method is
Estimating the Electoral Consequences of Legislative Redistricting
Abstract We analyze the effects of redistricting as revealed in the votes received by the Democratic and Republican candidates for state legislature. We develop measures of partisan bias and the
Why are American Presidential Election Campaign Polls so Variable When Votes are so Predictable
As most political scientists know, the outcome of the U.S. Presidential election can be predicted within a few percentage points (in the popular vote), based on information available months before
A Unified Method of Evaluating Electoral Systems and Redistricting Plans
We derive a unified statistical method with which one can produce substantially improved definitions and estimates of almost any feature of two-party electoral systems that can be defined based on
Why Are American Presidential Election Campaign Polls So Variable When Votes Are So Predictable ?
As most political scientists know, the outcome of the American presidential election can be predicted within a few percentage points (in the popular vote), based on information available months
...
...