Building a data foundation for DEIA
Without question, diverse perspectives will lead to greater scientific advancement and the acceleration of new breakthroughs. Many societies and publishers are actively working to support research initiatives focused on Diversity, Equity, Inclusion, and Accessibility (DEIA) within the industry, but there's still more work that needs to be done.
Without data, we can't understand the diversity of publishing programs or member communities. This same data can then be leveraged to foster a more diverse pool of peer reviewer pools, which is critical for equitable and inclusive scholarly evaluation. Additionally, data can highlight the vulnerabilities faced by marginalized groups when confronting emerging research integrity threats like paper mills. This understanding is pivotal in shaping our ability to provide the necessary support and mentorship.
Data-driven approaches effectively achieve two very important goals: the expansion of more diverse peer reviewer pools and the tracking of vulnerability among different groups when confronting emerging integrity threats.
Supercharging equity with plagiarism detection
In 2018, Retraction Watch found that the number of retractions due to plagiarism had more than tripled in the last twenty years, with a significant spike following the introduction of iThenticate in 2004.
Plagiarism rates were higher in articles with authors from low-income and non-English speaking countries. Many things could contribute to this increase. Researchers in the Global South may have less publishing experience, and need support to understand citation expectations. Their English-language skills may also make it harder to rephrase and recontextualize citable material. Rather than rejecting or judging these submissions, we as an industry need to provide the tools so all researchers can succeed.
But here's the beacon of hope: technology can step in to bolster plagiarism detection. By enhancing our ability to spot and prevent plagiarism, we can uphold integrity and level the playing field for these marginalized researchers, ensuring that they have access to tools that can help them make sure that their work is authentic.
Unbiasing with AI
How can we harness the potential of AI to promote equity?
While concerns about the increasing global use of AI are valid and growing, these tools also hold great power and potential. And, frankly, with the tools openly available to the world’s researchers, it's too late to put the cat back in the bag.
AI-powered tools, such as speech recognition and text-to-speech technologies, can enhance accessibility in research for individuals with disabilities, ensuring that their voices and perspectives are heard within the research community. AI tools can also aid researchers in pinpointing biases and inequalities within existing datasets, helping them detect more systemic issues that demand attention faster.
But what’s arguably the most compelling advantage of artificial intelligence for research equity is the potential of these tools to mitigate bias in peer review. When properly trained, AI has the power to reduce bias in peer review through language tools and support for researchers for whom English is not their native language. This can make things fairer and for scientists of all backgrounds, helping them focus on their work while AI helps them share their findings.
