viernes, 14 de junio de 2019

Taking – or leaving - a P: lifting a leg, or getting a leg-up on metrics of statistical rigor - On Medicine

Taking – or leaving - a P: lifting a leg, or getting a leg-up on metrics of statistical rigor - On Medicine



Christina Cantrell & James Giordano

Christina Cantrell, MS, RN, is an oncologic and critical and palliative care nurse, and research scholar in the Department of Physiology and Neuroethics Studies Program/Pellegrino Center for Clinical Bioethics at Georgetown University Medical Center, Washington, DC, USA.Her career interests are in the development and use of research methods to sustain and advance multi-disciplinary methods and applications in biomedicine and biotechnology in and across global communities.

James Giordano, PhD, MPhil is Professor in the Departments of Neurology and Biochemistry, and Chief of the Neuroethics Studies Program/Pellegrino Center for Clinical Bioethics, Georgetown University Medical Center, Washington, DC.As well, he is the Chair of the Neuroethics Program of the IEEE Brain Project, and has served as Senior Research Fellow of the EU-Human Brain Project Sub-project in Ethics. His ongoing research focuses upon processes and treatments of neuropsychiatric spectrum disorders, and the use of emerging techniques and technologies in research, and the applications of such methods and tools in medicine, public life, and global security. He is the Editor-in-Chief of the BMC/Springer journal Philosophy, Ethics, and Humanities in Medicine.


Taking – or leaving – a P: lifting a leg, or getting a leg-up on metrics of statistical rigor

Is it the end of the road for the p-value as we know it? Christina Cantrell and James Giordano, Editor in Chief of Philosophy, Ethics, and Humanities in Medicinetalk us through the recent discourse over the value of the p-value.
Recently Ioannidis et al. have proposed that a p value ≤ 0.05 may be an inadequate threshold to establish the statistical rigor needed to demonstrate outcomes’ validity. There may be something to this position. Increasing the stringency of p-values from ≤0.05 to ≤ 0.005 or 0.001 would certainly enable a more granular approach to – and criteria for – statistical significance.

No hay comentarios:

Publicar un comentario