scholarly journals Moving beyond noninformative priors: why and how to choose weakly informative priors in Bayesian analyses

Oikos ◽  
2019 ◽  
Vol 128 (7) ◽  
pp. 912-928 ◽  
Author(s):  
Nathan P. Lemoine
Keyword(s):  
Informative Priors ◽  
Bayesian Analyses ◽  
Noninformative Priors

Genotype-Guided P2Y 12 Inhibitor Therapy After Percutaneous Coronary Intervention: A Bayesian Analysi

2021 ◽  
Author(s):  
Vibhu Parcha ◽  
Brittain F. Heindl ◽  
Peng Li ◽  
Rajat Kalra ◽  
Nita A. Limdi ◽  
...  
Keyword(s):  
Myocardial Infarction ◽  
Percutaneous Coronary Intervention ◽  
Posterior Probability ◽  
Coronary Intervention ◽  
Cardiovascular Death ◽  
Minor Bleeding ◽  
Informative Priors ◽  
Noninformative Priors ◽  
Percutaneous Coronary ◽  
Guided Therapy

Background: Among patients receiving percutaneous coronary intervention (PCI), the role of a genotype-guided approach for antiplatelet therapy compared with usual care is unclear. We conducted a Bayesian analysis of the entire TAILOR-PCI (Tailored Antiplatelet Initiation to Lessen Outcomes Due to Decreased Clopidogrel Response After Percutaneous Coronary Intervention) randomized clinical trial population to evaluate the effect of the genotype-guided antiplatelet therapy post-PCI compared with the usual care on the risk of major adverse cardiovascular events (MACE). Methods: The primary outcome for our study was the composite of MACE (myocardial infarction, stroke, and cardiovascular death). Secondary outcomes included cardiovascular death, stroke, myocardial infarction, stent thrombosis, and major/minor bleeding. Bayesian modeling was used to estimate the probability of clinical benefit of genotype-guided therapy using (1) noninformative priors (ie, analyzing the TAILOR-PCI trial) and (2) informative priors derived from the ADAPT, POPular Genetics, IAC-PCI, and PHARMCLO trials (ie, analyzing TAILOR-PCI trial in the context of prior evidence). Risk ratio (RR: ratio of cumulative outcome incidence between genotype-guided and conventional therapy group) and 95% credible interval (CrI) were estimated for the study outcomes, and probability estimates for RR <1 were computed. Results: Using noninformative priors, in TAILOR-PCI the RR for MACE was 0.78 (95% CrI, 0.55–1.07) in genotype-guided therapy after PCI, and the probability of RR <1 was 94%. Using noninformative priors, the probability of RR <1 for cardiovascular death (RR, 0.95 [95% CrI, 0.52–1.74]), stroke (RR, 0.68 [95% CrI, 0.44–1.06]), myocardial infarction (RR, 0.84 [95% CrI, 0.37–1.89]), stent thrombosis (RR, 0.75 [95% CrI, 0.37–1.45]), and major or minor bleeding (RR, 1.22 [95% CrI, 0.84–1.77]) were 57%, 96%, 67%, 94%, and 15%, respectively. Using informative priors, the posterior probability of RR <1 for MACE, from genotype-guided therapy, was 99% (RR, 0.69 [95% CrI, 0.57–0.84]). Using informative priors, the posterior probability of RR <1 for cardiovascular death (RR, 0.86 [95% CrI, 0.61–1.19]), stroke (RR, 0.69 [95% CrI, 0.48–0.99]), myocardial infarction (RR:0.56 [95% CrI, 0.40–0.78]), stent thrombosis (RR, 0.59 [95% CrI, 0.38–0.94]), and major or minor bleeding (RR, 0.84 [95% CrI, 0.70–0.99]) were 81%, 99%, 99%, 99%, and 99%, respectively. Conclusions: Bayesian analysis of the TAILOR-PCI trial provides clinically meaningful data on the posterior probability of reducing MACE using genotype-guided P2Y 12 inhibitor therapy after PCI.

scholarly journals Bayesian Estimation with Informative Priors is Indistinguishable from Data Falsification

2019 ◽  
Vol 22 ◽  
Author(s):  
Miguel Ángel García-Pérez
Keyword(s):  
Bayesian Estimation ◽  
Likelihood Function ◽  
Research Integrity ◽  
Statistical Characteristics ◽  
Informative Priors ◽  
Bayesian Analyses ◽  
Interval Estimates ◽  
Frequentist Statistics ◽  
Qualitative Level ◽  
Strong Inference

Abstract Criticism of null hypothesis significance testing, confidence intervals, and frequentist statistics in general has evolved into advocacy of Bayesian analyses with informative priors for strong inference. This paper shows that Bayesian analysis with informative priors is formally equivalent to data falsification because the information carried by the prior can be expressed as the addition of fabricated observations whose statistical characteristics are determined by the parameters of the prior. This property of informative priors makes clear that only the use of non-informative, uniform priors in all types of Bayesian analyses is compatible with standards of research integrity. At the same time, though, Bayesian estimation with uniform priors yields point and interval estimates that are identical or nearly identical to those obtained with frequentist methods. At a qualitative level, frequentist and Bayesian outcomes have different interpretations but they are interchangeable when uniform priors are used. Yet, Bayesian interpretations require either the assumption that population parameters are random variables (which they are not) or an explicit acknowledgment that the posterior distribution (which is thus identical to the likelihood function except for a scale factor) only expresses the researcher’s beliefs and not any information about the parameter of concern.

Performance of informative priors skeptical of large treatment effects in clinical trials: A simulation study

2015 ◽  
Vol 27 (1) ◽  
pp. 79-96 ◽  
Author(s):  
Claudia Pedroza ◽  
Weilu Han ◽  
Van Thi Thanh Truong ◽  
Charles Green ◽  
Jon E Tyson
Keyword(s):  
Clinical Trials ◽  
Simulation Study ◽  
Treatment Effects ◽  
Mean Squared Error ◽  
Binary Outcomes ◽  
Control Group ◽  
Informative Priors ◽  
Bayesian Analyses ◽  
Credible Intervals ◽  
True Treatment Effect

One of the main advantages of Bayesian analyses of clinical trials is their ability to formally incorporate skepticism about large treatment effects through the use of informative priors. We conducted a simulation study to assess the performance of informative normal, Student- t, and beta distributions in estimating relative risk (RR) or odds ratio (OR) for binary outcomes. Simulation scenarios varied the prior standard deviation (SD; level of skepticism of large treatment effects), outcome rate in the control group, true treatment effect, and sample size. We compared the priors with regards to bias, mean squared error (MSE), and coverage of 95% credible intervals. Simulation results show that the prior SD influenced the posterior to a greater degree than the particular distributional form of the prior. For RR, priors with a 95% interval of 0.50–2.0 performed well in terms of bias, MSE, and coverage under most scenarios. For OR, priors with a wider 95% interval of 0.23–4.35 had good performance. We recommend the use of informative priors that exclude implausibly large treatment effects in analyses of clinical trials, particularly for major outcomes such as mortality.

Bayesian analyses of cognitive architecture.

2017 ◽  
Vol 22 (2) ◽  
pp. 288-303 ◽  
Author(s):  
Joseph W. Houpt ◽  
Andrew Heathcote ◽  
Ami Eidels
Keyword(s):  
Cognitive Architecture ◽  
Bayesian Analyses

No Evidence that Experiencing Physical Warmth Promotes Interpersonal Warmth: Two Failures to Replicate Williams and Bargh (2008)

2018 ◽  
Author(s):  
Christopher Chabris ◽  
Patrick Ryan Heck ◽  
Jaclyn Mandart ◽  
Daniel Jacob Benjamin ◽  
Daniel J. Simons
Keyword(s):  
Null Hypothesis ◽  
Small Sample ◽  
Sample Sizes ◽  
Double Blind ◽  
Bayesian Analyses ◽  
Physical Warmth ◽  
Small Sample Sizes ◽  
Interpersonal Warmth

Williams and Bargh (2008) reported that holding a hot cup of coffee caused participants to judge a person’s personality as warmer, and that holding a therapeutic heat pad caused participants to choose rewards for other people rather than for themselves. These experiments featured large effects (r = .28 and .31), small sample sizes (41 and 53 participants), and barely statistically significant results. We attempted to replicate both experiments in field settings with more than triple the sample sizes (128 and 177) and double-blind procedures, but found near-zero effects (r = –.03 and .02). In both cases, Bayesian analyses suggest there is substantially more evidence for the null hypothesis of no effect than for the original physical warmth priming hypothesis.

Sign in / Sign up

Export Citation Format

Share Document