Andrew Balas from Augusta University said that researchers accomplish a lot but they also have many responsibilities. They have to separate what is good from what is not so good. Major quality defects in research include non-reproducible results, deficiency in design and conduct deficiencies, uninformative results, and unfinished or non-reported clinical trials. In a well known incident, the president of Stanford University resigned after a report issued in 2013 found manipulated data and a failure to correct mistakes in his research. There is a constant pressure to produce more and more; Peter Higgs who won the 2013 Nobel Prize in physics said that he would not take an academic job today because he might not be regarded as productive enough. Sometimes failures are not clear; 97% of oncology drug pairs tested in clinical trials never receive FDA approval. Huge benefits of research come from mostly wasted effort!  Publishing needs to change and embrace quality control. To eliminate binary defects of controlled clinical trials, Lancet now requires more than one author of an article to have direct access to the raw data so it can be verified.

Gwen Evans from Elsevier presented the publisher perspectives on research quality and research integrity. AI generated content is everywhere. Fraudulent research is on the rise, particularly in STEM fields because of the growth of research, pressure to publish,  the rise of Gen AI, and the scholarly communication landscape getting more complex. Elsevier now has around 16,000 articles under investigation by its Research Integrity team, and it estimates that it will retract 1,500 to 2,000 articles this year. The rise of Gen AI is a new challenge.

Fraud is a business for paper mills, so Elsevier is staffing for research integrity. Human oversight and a lot of care are required so as not to ruin people's lives and careers. 

Elsevier has developed a pre-publication checker. Libraries also have a role to play. Lack of confidence in the research process is a growing concern; for example the Wall Street Journal published articles calling for the firing of a scientist who claimed to have discovered a room temperature superconductor.

Brad Warren, also from Augusta University, listed some library concerns: quality, promotion and tenure, editorial and peer review. Future concerns include AI and data mining, development of standards, and threats to academic freedom. If research is faulty, why do we put so much effort into collecting and preserving it? Many tools are based on usage which is like tinder; what is being done with this? Tools are based on prestige, but where does prestige derive from? Is it the peer review process or something else? Teaching and research are always more highly regarded than service. Why is the service piece not as important as the research? How does this change our approach to scholarly communication? The industry moves fast, but editorial boards meet less frequently. What do we do with people who continue to submit shoddy articles? We should focus on continued access.

Donald T. Hawkins, Conference Blogger