Limits...
Model Averaging for Improving Inference from Causal Diagrams.

Hamra GB, Kaufman JS, Vahratian A - Int J Environ Res Public Health (2015)

Bottom Line: We use three techniques for averaging the results among multiple candidate models: information criteria weighting, inverse variance weighting, and bootstrapping.We show that each averaging technique returns similar, model averaged causal estimates.An a priori strategy of model averaging provides a means of integrating uncertainty in selection among candidate, causal models, while also avoiding the temptation to report the most attractive estimate from a suite of equally valid alternatives.

View Article: PubMed Central - PubMed

Affiliation: Department of Environmental and Occupational Health, Drexel University School of Public Health, Philadelphia, PA 19104, USA. ghassan.b.hamra@drexel.edu.

ABSTRACT
Model selection is an integral, yet contentious, component of epidemiologic research. Unfortunately, there remains no consensus on how to identify a single, best model among multiple candidate models. Researchers may be prone to selecting the model that best supports their a priori, preferred result; a phenomenon referred to as "wish bias". Directed acyclic graphs (DAGs), based on background causal and substantive knowledge, are a useful tool for specifying a subset of adjustment variables to obtain a causal effect estimate. In many cases, however, a DAG will support multiple, sufficient or minimally-sufficient adjustment sets. Even though all of these may theoretically produce unbiased effect estimates they may, in practice, yield somewhat distinct values, and the need to select between these models once again makes the research enterprise vulnerable to wish bias. In this work, we suggest combining adjustment sets with model averaging techniques to obtain causal estimates based on multiple, theoretically-unbiased models. We use three techniques for averaging the results among multiple candidate models: information criteria weighting, inverse variance weighting, and bootstrapping. We illustrate these approaches with an example from the Pregnancy, Infection, and Nutrition (PIN) study. We show that each averaging technique returns similar, model averaged causal estimates. An a priori strategy of model averaging provides a means of integrating uncertainty in selection among candidate, causal models, while also avoiding the temptation to report the most attractive estimate from a suite of equally valid alternatives.

No MeSH data available.


Related in: MedlinePlus

Directed Acyclic graph to obtain an unbiased effect of pre-pregnancy weight on cesarean delivery; adapted from Vahratian et al. (2005). Sufficient sets from this DAG are determined using DAGGITY software.
© Copyright Policy
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC4555287&req=5

ijerph-12-09391-f002: Directed Acyclic graph to obtain an unbiased effect of pre-pregnancy weight on cesarean delivery; adapted from Vahratian et al. (2005). Sufficient sets from this DAG are determined using DAGGITY software.

Mentions: The authors of the original article provided a DAG summarizing the potential confounders of interest in their analyses. A group of maternal characteristics were placed within a single node of the DAG; this provided a streamlined presentation, but did not allow visualization of the relationships of each these variables to others, and each other. To facilitate determination of sufficient adjustment sets, we disaggregated these variables so each has its own node, and we added arrows for the relationships of these variables to others in the DAG. Further, to aid in visualization, we remove variables that were in the original DAG but would clearly not be considered in any minimally sufficient adjustment set. Our modified DAG is presented in Figure 2.


Model Averaging for Improving Inference from Causal Diagrams.

Hamra GB, Kaufman JS, Vahratian A - Int J Environ Res Public Health (2015)

Directed Acyclic graph to obtain an unbiased effect of pre-pregnancy weight on cesarean delivery; adapted from Vahratian et al. (2005). Sufficient sets from this DAG are determined using DAGGITY software.
© Copyright Policy
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC4555287&req=5

ijerph-12-09391-f002: Directed Acyclic graph to obtain an unbiased effect of pre-pregnancy weight on cesarean delivery; adapted from Vahratian et al. (2005). Sufficient sets from this DAG are determined using DAGGITY software.
Mentions: The authors of the original article provided a DAG summarizing the potential confounders of interest in their analyses. A group of maternal characteristics were placed within a single node of the DAG; this provided a streamlined presentation, but did not allow visualization of the relationships of each these variables to others, and each other. To facilitate determination of sufficient adjustment sets, we disaggregated these variables so each has its own node, and we added arrows for the relationships of these variables to others in the DAG. Further, to aid in visualization, we remove variables that were in the original DAG but would clearly not be considered in any minimally sufficient adjustment set. Our modified DAG is presented in Figure 2.

Bottom Line: We use three techniques for averaging the results among multiple candidate models: information criteria weighting, inverse variance weighting, and bootstrapping.We show that each averaging technique returns similar, model averaged causal estimates.An a priori strategy of model averaging provides a means of integrating uncertainty in selection among candidate, causal models, while also avoiding the temptation to report the most attractive estimate from a suite of equally valid alternatives.

View Article: PubMed Central - PubMed

Affiliation: Department of Environmental and Occupational Health, Drexel University School of Public Health, Philadelphia, PA 19104, USA. ghassan.b.hamra@drexel.edu.

ABSTRACT
Model selection is an integral, yet contentious, component of epidemiologic research. Unfortunately, there remains no consensus on how to identify a single, best model among multiple candidate models. Researchers may be prone to selecting the model that best supports their a priori, preferred result; a phenomenon referred to as "wish bias". Directed acyclic graphs (DAGs), based on background causal and substantive knowledge, are a useful tool for specifying a subset of adjustment variables to obtain a causal effect estimate. In many cases, however, a DAG will support multiple, sufficient or minimally-sufficient adjustment sets. Even though all of these may theoretically produce unbiased effect estimates they may, in practice, yield somewhat distinct values, and the need to select between these models once again makes the research enterprise vulnerable to wish bias. In this work, we suggest combining adjustment sets with model averaging techniques to obtain causal estimates based on multiple, theoretically-unbiased models. We use three techniques for averaging the results among multiple candidate models: information criteria weighting, inverse variance weighting, and bootstrapping. We illustrate these approaches with an example from the Pregnancy, Infection, and Nutrition (PIN) study. We show that each averaging technique returns similar, model averaged causal estimates. An a priori strategy of model averaging provides a means of integrating uncertainty in selection among candidate, causal models, while also avoiding the temptation to report the most attractive estimate from a suite of equally valid alternatives.

No MeSH data available.


Related in: MedlinePlus