As you look for and evaluate scholarly sources, you may see the ProQuest Research Assistant, which is an AI tool. A few things to consider.
The ProQuest Research Assistant uses a Large Language Mode (LLM)/Chat GPT on the data from the database(s).
LLMs are trained on vast amounts of data to predict the next word in a sentence. The responses can be helpful but AI tools don't know your research context, what you have read, what your knowledge of the topic is, etc.
Even if the LLM draws on "good data sources," it still has risks of inaccuracies and hallucination. Verify any citation generated by the AI tool; use the Library Search tool or contact me.
For a document's "Key Takeaways," the ProQuest Research Assistant uses the same perspective as the original document. Therefore, it might not note overlaps, contradictions, agreements, etc. with other scholarly sources. Research does not happen in a vacuum but is part of a larger "conversation" where authors debate, refute, support, and extend research, theories, and claims. Consider other perspectives and voices in your search/evaluation of a scholarly source; develop your own "key takeaways." To a novice or developing researcher, AI can generate output that looks legitimate, but the tool can miss little things that only a seasoned researcher would notice.
See more about ProQuest Research Assistant
Sources consulted:
Coursera courses offered by Jules White: Generative AI Primer, Trustworthy Generative AI
Sociology 3242: AI and Social Systems syllabus, courtesy of Vanderbilt Professor Jenny Davis
