Multiple‐bias modelling for analysis of observational data

By Sander Greenland. 2005

497 citations recorded by the source corpus.

View Paper: https://www.semanticscholar.org/paper/bf7519b9d071643d6111d24744520d77efb9018d

How researchers see this paper

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

In Statistics and Probability

Citations in the corpus

What Are the Limits of Posterior Distributions Arising From Nonidentified Models, and Why Should We Care? (2009, background): “(For a general discussion of this point, see Greenland 2003, 2005.)”
GetReal in network meta‐analysis: a review of the methodology (2016, methodology): “More advanced methods account for multiple sources of bias, where the contribution of each source is modeled separately (see, e.g., Eddy and Shachter, 1992; Greenland, 2003, 2005; Wolpert and Mengersen, 2004).”
Relaxation Penalties and Priors for Plausible Modeling of Nonidentified Bias Sources (2009, methodology): “With a noninformative prior for EXY and p(βT |EXY ) = p(βT ), as here, step (2) reduces to drawing β∗T from p(βT ) and the resulting sampler is approximated by Monte Carlo Sensitivity Analysis (MCSA) in which bootstrap draws a∗XY from aXY replace E∗XY (Greenland, 2005a).”
Transparent Parametrizations of Models for Potential Outcomes (2012, methodology): “Following Leamer (1978), Gustafson (2005) and Greenland (2005), we argue that partially identified models should be re-parameterized so 570 T. S. Richardson, R. J. Evans and J. M. Robins that the complete parameter vector may be divided into point-identified and entirely non-identified subvectors.”
Meta-analysis for linear and nonlinear dose-response relations: examples, an evaluation of approximations, and software. (2012, background): “We did not consider other important issues, such as optimal choices of dose values for exposure levels, publication bias, and methodological bias (6, 10, 18).”
A re-evaluation of random-effects meta-analysis (2009, background): “Alternative approaches to dealing with specific sources of bias are available (Eddy et al., 1992; Greenland, 2005).”
Causal Effects In Psychotherapy: Counterfactuals Counteract Overgeneralization (2010, methodology): “Some of these methods (Bayesian bias modeling and Monte Carlo sensitivity analysis; Greenland 2005b; Rothman et al., 2008, chapter 9) take the uncertainty in so-called bias parameters into account to yield a more realistic impression of the uncertainty in the EM estimate.”
Local model uncertainty and incomplete‐data bias (with discussion) (2005, background): “Greenland (2005), section 4.5, discusses the same issue, emphasizing the inadequacy of methods which claim to ‘let the data speak for themselves’, pointing out that ‘without external inputs, observational data say nothing at all about causal effects’.”
On Instrumental Variables Estimation of Causal Odds Ratios (2011, background): “Concerns about the validity of this assumption plague observational data analyses and increase the uncertainty surrounding many study results (Greenland, 2005).”
Facilitating meta‐analyses by deriving relative effect and precision estimates for alternative comparisons from a set of estimates presented by exposure level or disease category (2008, background): “The table allows adjusted effect estimates and CIs to be calculated for any alternative comparison of levels (including a dose-related trend statistic), and can help a dose–response meta-analysis and/or support a sensitivity analysis for methodological bias [16].”
The Bradford Hill considerations on causality: a counterfactual perspective (2005, unknown): “Contemporary methods for multiple bias modelling include Bayesian methods and Monte Carlo sensitivity analysis (MCSA, which can be modified to be approximatively interpretable in a Bayesian manner under certain conditions [27]).”
Good practices for quantitative bias analysis. (2014, methodology): “Many bias analysis methods assume no effectmeasure modification by the unmeasured confounder, although methods to account for effect-measure modification are available.(15,28) A mirror problem to uncontrolled confounding is overadjustment bias, which arises from adjustment for inappropriate variable”