Conectar
To demonstrate a worked-out example of a Bayesian independent t-test using open-source software, simulated data, a hypothetical nurse education intervention and a randomised controlled study design. This tutorial explains relevant Bayesian concepts and highlights literature that provides statistically principled justifications for replacing or complementing the frequentist independent t-test with its Bayesian counterpart.
Bayesian t-test analysis tutorial.
A pedagogical framework was applied.
Simulated data generated in Microsoft Excel was uploaded to the Open Science Framework, accessible at: osf.io/4t9gn.
The Bayesian independent t-test in JASP provides: (1) a Bayes factor quantifying the relative evidence for determining which of two competing theories, that is, the null (H0) or the alternative (H1) hypotheses, better supports the experimental data and (2) the posterior probability distribution, with its median point estimate plus a 95% credible interval, quantifying the magnitude and uncertainty of the effect size estimate.
This article provides a practical method for nursing and midwifery researchers to conduct Bayesian analysis, offering statistical, practical and ethical advantages, including the application of sequential analysis and optimal stopping rules enhancing research efficiency.
This article increases awareness of the feasibility and benefits of Bayesian analysis in nursing and midwifery research, emphasising its ease of implementation through open-source software. Clear step-by-step guidance is provided to support its wider adoption and strengthen methodological rigour in nursing and midwifery research.
Nursing and midwifery research has traditionally relied upon frequentist statistical techniques, based on p values and confidence intervals. Bayesian methods can: (a) improve nursing and midwifery decision-making with probabilistic evidence and (b) reduce publication bias by avoiding binary interpretation of research results.
The methodology aligns with van Doorn et al. (2021) guidelines for conducting and reporting a Bayesian analysis.
No patient or public contribution.
FreshRSS