AI in Research and Scholarship
Consults and Solutions
Learn about using Generative AI in your research, access to technology, and potential AI applications. Send us an email at researchalliance@montana.edu.
Access
Find out about access to generative AI tools, custom models that can help your research, and how to start using generative AI for your research.
Training
Talk through existing training, work through custom training needs, and learn more about the types of training that can help you.
About
A partnership between the MSU Library, Research CyberInfrastructure, and the Center for Faculty Excellence to address questions on Generative AI in research, technology access, and more.
Thinking about AI in your teaching? Visit the Teaching & Learning information on generative AI.
Howto Integrate Generative AI into Your Research and Scholarship
Overview
As generative AI is being increasingly integrated into scholarly work, including graduate theses and doctoral dissertations, and journal articles, the following guidance can help faculty and graduate students with ethical and responsible use of AI in their scholarly work.
To maintain the highest standards of academic quality and research integrity, graduate students and faculty advisors should make every attempt to practice full transparency. In achieving transparency, graduate students should declare how they plan to use generative AI to their faculty advisors to get approval for this use. Following good practices for other scholarly writing and disciplinary norms, students should cite when and describe how generative AI tools were used so that audiences of their writing can distinguish between the contributions of the writer and the generative AI tool. This, for example, may include citing and describing use of generative AI tools in searching, designing, outlining, drafting, writing, or editing the thesis, or in producing audio or visual content for the thesis, and may include other uses of generative AI.
Note that departments and colleges may have different or additional specific requirements or restrictions on the use of generative AI during the phases of the research and writing cycle that correspond to their research methods and analytical processes. This could include, for example, guidance on use in writing text, conducting analytical work, reporting results (e.g., tables or figures) or writing computer code. To adhere to potential additional departmental requirements, check with your advisor and/or committee.
Yes, generative AI tools and techniques can be integrated into your research. However, navigating the ethical and methodological considerations of incorporating generative AI into research requires careful planning and thoughtful implementation. Read on for a series of potential steps and guidelines to consider as you look to incorporate these technologies into your research process.
Step 1: Seek Guidance from Your Advisor or Editor
Before starting your research, it is crucial to seek the guidance of your advisor
or editor. Their expertise and experience will prove invaluable as you assess the
suitability of AI for your research question or scholarship, identify potential ethical
concerns, and ensure compliance with institutional policies. Open communication will
lay the foundation for a successful and ethically sound research endeavor.
Step 2: Identify the Right AI Tools and Techniques
Generative AI encompasses a diverse spectrum of tools and techniques, each with its
unique strengths and limitations. Carefully evaluate your research objectives and
data characteristics to select the AI tools that best align with your needs. Consult
with experts in the field and explore relevant literature to gain a comprehensive
understanding of the available AI options. Through discussions with your advisor or
editor, declare how you will be using the technology.
Step 3: Gather and Prepare Your Data
The quality of your data is paramount to the success of AI-driven research. Ensure
that your data is accurate, consistent, and free from biases. Employ appropriate data
cleaning and preprocessing techniques to prepare your data for analysis by AI algorithms.
Step 4: Train and Validate Your AI Models
Developing and refining AI models is an iterative process. Carefully select the training
data and parameters to ensure that your models are robust and generalizable. Employ
rigorous validation techniques to assess the performance of your models.
Step 5: Interpret and Communicate Your Results
AI algorithms can produce complex and sometimes counterintuitive results. Critically
evaluate the outputs of your AI models and usage. As questions arise, seek expert
guidance to ensure that your interpretations are sound and supported by evidence.
Clearly communicate your findings in a manner that is accessible to both technical
and non-technical audiences.
Step 6: Cite AI Resources Appropriately
Just as you would cite traditional research sources, it is essential to properly acknowledge
the AI tools and resources that you have utilized. Follow established citation guidelines
to ensure that your work is transparent and reproducible. (See section below on Describing
and referencing generative AI tools and usage.)
Step 7: Address Ethical Considerations
AI applications raise a range of ethical concerns, including data privacy, biased
representations, fairness, and transparency. Carefully consider these implications
throughout your research process and implement appropriate safeguards to mitigate
potential risks.
Step 8: Document Your AI Methodology
Provide detailed documentation of your AI methodology, including the specific tools,
algorithms, and parameters employed. This documentation will facilitate reproducibility
and allow others to assess the validity of your findings.
There are several key reasons why generative AI tools and techniques should be appropriately cited when used within your research and scholarship:
- Transparency in your research methods is crucial for evaluating the quality and reproducibility of findings. Clearly attributing any use of generative AI provides transparency and necessary context. Moreover, citing implementation details enables accurate interpretation.
- If generative AI systems significantly contribute to your research, they should receive due credit like any other contributor. Citing AI recognizes valuable technological assistance and the hybrid nature of human/machine collaborations.
- Documenting AI use establishes provenance regarding generated content and accountability for outputs. Cataloging AI adoption through citation practices provides a record of its growing influence and patterns of substitution for human effort over time. Citations of AI tools and applications create valuable records which indicate and benchmark evolving technology use.
Generative AI Citation Examples – Multiple Styles
APA style
Currently, APA recommends that text generated from AI be formatted as "Personal Communication."
As such, it receives an in-text citation but not an entry on the References list.
Rule: (Communicator, personal communication, Month Date, Year)
Examples:
(OpenAI, personal communication, January 16, 2023).
When asked to explain psychology's main schools of thought, OpenAI's ChatGPT's response
included ... (personal communication, February 22, 2023).
MLA style
The Modern Language Association provides detailed guidance on citing generative AI
according to their template.
Chicago style
Chicago Style with footnotes
Personal communications are cited in a footnote, but are not listed in the bibliography.
Rule: Number.Originator of the communication, medium, Day Month, Year.
Example:
OpenAI's ChatGPT AI language model, response to question from author, 7 February,
2023.
Shortened note rule: Number, Correspondent's last name, medium
Example:
ChatGPT, response to prompt from author
Even when engaging in authorized generative AI use, faculty and graduate students must be aware of the risks in using such tools, some of which are discussed below.
Privacy and Secured Data
It is unclear how companies that host and support generative AI use the input (prompts) and output data from these tools. This can potentially raise privacy and security concerns when using generative AI tools to analyze sensitive data collected for human-based research or a company’s proprietary technology. If a student or faculty plans to use generative AI tools to analyze data, they must include this use for approval in any applications for the institutional review board (IRB).
Inaccurate and Biased Content
Generative AI is prone to producing content that blends facts with false and biased information to make it appear true or valid. AI tools can reproduce biases that exist in the content it is trained on, including presenting untrue statements as facts the perpetuate biases from offensive content that discriminates against marginalized groups based on gender, race, and sexual orientation. Generative AI can at times make up references to scholarly work that does not exist. Students and faculty are ultimately responsible for all the content of their scholarly work, including content generated by AI tools.
Note that Generative AI tools predict based on existing content meaning it may not be as capable of producing, or rearranging existing knowledge in ways to achieve, the type of new and novel content that meets the expectations and standards of research across fields and disciplines.
Relevant Montana State University (MSU) Policies
- MSU Dean of Students Plagiarism Policy?
- Other local policies or guidance?
- Other external policies
Further Reading and Viewing
- Mohammad Hosseini, Lisa M. Rasmussen & David B. Resnik (2023). Using AI to write scholarly publications. Accountability in Research, DOI: 10.1080/08989621.2023.2168535
- Nature (2023). Policy on Artificial Intelligence in Journal Submissions. [Online]. Available: https://www.nature.com/nature-portfolio/editorial-policies/ai [Accessed 4 December 2023].
- White House Office of Science and Technology Policy. Blueprint for an AI Bill of Rights: Making Automated Systems Work for the American People. [Online]. Available: https://www.whitehouse.gov/ostp/ai-bill-of-rights/ [Accessed 04 December 2023].
Works Cited
The School of Graduate Studies at the University of Toronto. Guidance on the Appropriate Use of Generative Artificial Intelligence in Graduate Theses. Available: https://www.sgs.utoronto.ca/about/guidance-on-the-use-of-generative-artificial-intelligence/ [Accessed October 3, 2023].