Diagnostic Checks for Multilevel Models

By Tom A. B. Snijders, Johannes Berkhof. 2008

161 citations recorded by the source corpus.

Subfields: Genetics, Statistics and Probability

View Paper: https://www.semanticscholar.org/paper/f78d7e8d9e612263644e653785b55df3a0533049

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Genetics

In Statistics and Probability

Citations in the corpus

Best-Practice Recommendations for Defining, Identifying, and Handling Outliers (2013, background): “First, the recommendation is to calculate the average squared deviation, or Cj, for each group of cases, where the focal group is denoted by j (Snijders & Berkhof, 2008; Snijders & Bosker, 2012).”
The Role of Sample Cluster Means in Multilevel Models (2011, background): “This issue is well known in the setting of panel data (Hausman and Taylor, 1981), but the topic has recently received attention also in a more general perspective: see Skrondal and Rabe-Hesketh (2004), Fielding (2004), Ebbes et al. (2004), Kim and Frees (2007) and Snijders and Berkhof (2008).”
Aggregated, Voluntary and Mandatory Risk Disclosure Incentives: Evidence from UK FTSE All Share Companies (2013, methodology): “Both the Q-Q plot and the normality curve with histogram for standardised residuals and predicted values of ARD, VRD and MRD visually support the validation of the normality assumption (e.g., Snijders and Berkhof, 2008).”
Multilevel Modeling with Correlated Effects (2007, background): “Snijders and Berkhof (2007) discussed diagnostics for multilevel model specification.”
Multilevel and related models for longitudinal data (2008, background): “As discussed for linear models in Chapter 3 [61], violation of the assumption that δj has zero expectation can invalidate inference.”
Bayesian Multilevel Analysis and MCMC (2008, unknown): “[61] for an application of Laplace approximations”
Genetic association testing using the GENESIS R/Bioconductor package (2019, methodology): “In addition to the standard GRM/KM term, other variance components can be used to account for e.g. shared environments (Conomos et al., 2016b), as well as phenotypic heteroskedasticity by study subgroup (Snijders and Berkhof, 2008).”
influence.ME: Tools for Detecting Influential Data in Mixed Effects Models (2012, unknown): “A formula for Cook’s Distance is provided (Snijders and Bosker, 1999; Snijders and Berkhof, 2008): C0Fj = 1 r + 1 ( γ̂− γ̂(−j) )′ Σ̂−1F ( γ̂− γ̂(−j) ) in which γ̂ represents the vector of original parameter estimates, γ̂(−j)the parameter estimates of the model excluding higher-level unit j, and Σ̂F…”