This article is more than 5 years old. Information may no longer be current.
Effective interventions needed to improve quality, credibility of research
Click Here to Manage Email Alerts
VIENNA, Austria — Though the task force of research and the number of published papers have grown tremendously in recent years, effective interventions are needed to improve the quality and credibility of scientific investigation, according to John P.A. Ioannidis, MD, a keynote lecturer at the Advanced Retinal Therapy meeting.
Between 1996 and 2011, 15 million scientists contributed to the publication of 25 million papers in scientific journals, according to Ioannidis.
John P.A. Ioannidis
However, “true and readily applicable major discoveries are far fewer, and many published research findings are false or exaggerated,” Ioannidis said.
The typical recipe for research practices, he said, includes solo, siloed investigators, small sample size studies, cherry picking of the best hypotheses, minimum significance thresholds, no registration, no replication and no data sharing.
In order to make more published research true, Ioannidis advocated the adoption of large-scale, collaborative research with replication culture; registration of studies, protocols, analysis codes, datasets, raw data and results; and the sharing of data, protocols, materials and software.
Research should also adopt more appropriate statistical methods, standardized definitions and more stringent thresholds for the definition of success, according to Ioannidis. Improved study design standards would also lead to more reliable results.
Finally, he pointed out that research should contain the influence of conflicted stakeholders and authors and be transparent about conflicts of interest.
Disclosure: Ioannidis has no relevant financial disclosures.