Javascript must be enabled for the correct page display

A Comparison of the Bayesian and Frequentist Approaches to Equivalence, Non-Inferiority, and Superiority Designs

Linde, Maximilian (2019) A Comparison of the Bayesian and Frequentist Approaches to Equivalence, Non-Inferiority, and Superiority Designs. Research Project 2 (major thesis), Behavioural and Cognitive Neurosciences.

[img]
Preview
Text
mBCN_2019_LindeM.pdf

Download (1MB) | Preview
[img] Text
toestemming.pdf
Restricted to Registered users only

Download (121kB)

Abstract

Clinical trials often seek to determine the equivalence, non-inferiority, or superiority of an experimental condition (e.g., a new drug) compared to a control condition (e.g., a placebo or an already existing drug). The use of frequentist statistical methods to analyse data for these types of designs is widespread. Importantly, however, frequentist inference has several limitations. Bayesian statistics remedies these shortcomings and allows for intuitive interpretations. The goal of the present article is twofold. First, we present baymedr, an R package that provides user-friendly tools for the computation of Bayes factors for equivalence, non-inferiority, and superiority designs (see also van Ravenzwaaij, Monden, Tendeiro, & Ioannidis, 2019). Second, we conducted simulations to contrast the performances of the Bayesian and frequentist equivalence and non-inferiority tests. We generated data sets with various true population effect sizes and sample sizes, which were analysed by the frequentist and Bayesian equivalence and non-inferiority tests. The resulting p-values and Bayes factors formed the basis for subsequent receiver operating characteristic (ROC) analyses. The classification performances of the Bayesian tests were higher compared to their frequentist counterparts. Generally, the frequentist equivalence and non-inferiority tests demanded a high sample size to reach a proper power. Together with the theoretical advantages of Bayesian inference, this leads us to propose the adaptation of our state-of-the-art Bayesian tools for the analysis of equivalence, non-inferiority, and superiority studies.

Item Type: Thesis (Research Project 2 (major thesis))
Supervisor:
Supervisor nameSupervisor E mail
Ravenzwaaij, D. vanD.van.Ravenzwaaij@rug.nl
Degree programme: Behavioural and Cognitive Neurosciences
Thesis type: Research Project 2 (major thesis)
Language: English
Date Deposited: 26 Jul 2019
Last Modified: 29 Jul 2019 11:36
URI: http://fse.studenttheses.ub.rug.nl/id/eprint/20456

Actions (login required)

View Item View Item