Special Forum: AI through the ΑΩ: Theological Librarians Interact with Artificial Intelligence

What Does ChatGPT Have to Do with Library Instruction and Reference Interviews?

Do you remember the time before the World Wide Web and smartphones? Consider how much these two developments have impacted the world in the past thirty years. Artificial intelligence (AI) tools such as ChatGPT will undoubtedly have similar impacts in the years to come, and some of these changes raise concerns regarding academic integrity in higher education. Among them is the increased potential for student plagiarism, which is “the action or practice of taking someone else’s work, idea, etc., and passing it off as one’s own; literary theft” (Oxford English Dictionary, 2023).

In the past, educators could turn to plagiarism software to identify students’ unethical use of articles, websites, or papers written by others. The challenge with ChatGPT is that AI-generated papers are no longer based on any specific resource. Instead, these tools generate new content based on many sources. For example, ChatGPT, which is built on a series of generative pre-trained transformer (GPT) models, uses countless sources to “write” a new text that did not exist previously. As a result, the traditional way to detect plagiarism—checking if a student copied an existing work—is outdated. New generations of plagiarism checkers have emerged; however, such products have been unreliable. Udell states that ZeroGPT, an AI writing detector, determined with 88.2 percent certainty that Genesis, the first book of the Bible, was written by AI (2024, 23). While developers may continue to fine-tune the accuracy of tools such as ZeroGPT, it is also likely that generative-text AIs will remain a step ahead.

Suppose one accepts the definition of plagiarism quoted above—“literary theft.” In that case, submitting a paper that was largely or entirely composed by AI is unethical if viewed through the lenses of many religious traditions. Judeo-Christian faith traditions have well-known prohibitions against theft (Exod. 20:15, Mark 10:19). One could argue that using AI is not plagiarism in the strictest sense, because the user of AI is not copying any one text and not stealing from any particular author. However, a person who uses any generative AI program to compose a paper is nevertheless stealing. After all, a person who steals parts to build a TV is just as much a thief as someone who breaks into a home to steal an assembled TV.

How might libraries actively help individuals who are tempted to take shortcuts and use generative AI to write research papers or other assignments? How might academic libraries help faculty prevent students from taking such shortcuts? This author believes these questions are answered in traditional library instruction and reference interviews.

To prevent plagiarism, one must first understand why individuals, particularly students, would resort to copying materials or using AI. Hattingh, Buitendag, and Lall provide a list of such reasons: (1) some students do not understand what constitutes plagiarism or are unfamiliar with citation rules, (2) they may take shortcuts to receive a higher grade or reduce the amount of time to complete an assignment, (3) some plagiarize because they lack time management skills and are in danger of completing the assignment late, (4) students may be afraid of failure and see plagiarizing as the only way to accomplish the task, (5) they may plagiarize to rebel against authority or simply lack motivation to complete the task, (6) some students plagiarize because of their attitudes toward instructors, in that some think their teacher permits the practice while others believe their teacher’s requirements are too demanding, (7) some may believe that since their peers plagiarize, their own plagiarism is justified, (8) some students cannot resist the temptation to plagiarize when they see that the information is so easily accessible, (9) some perceive very little risk in getting caught or punished, (10) they may feel overwhelmed by the assignment, and (11) many experience internal and external pressures to perform well and see plagiarism as the only way to succeed academically (2020, 171–173). One could further add that (12) some students have a genuine lack of understanding of the research process (Paramasivam et al. 2023, 107; Šprajc et al. 2017, 42). In other words, some students do not know where or how to start and complete the research process.

While this list is probably not exhaustive, and many of these reasons may be outside a library’s sphere of influence (probably reasons 5–9), librarians can actively counter several motivations for plagiarism mentioned above. It is beyond the scope of this article to cite all possible solutions to these reasons. However, here are a few points that demonstrate opportunities for librarians to intervene through reference interviews or classroom instruction:

  • Teaching faculty and librarians should collaborate to provide in-class information literacy instructions to introduce and emphasize proper citation practices to prevent accidental plagiarism (reason 1) and make students aware of the potential penalties for plagiarism (reasons 6, 9). Instructors and librarians should further collaborate to make students aware of resources and services, such as a reference interview, available in the library.
  • Teaching faculty may require students to meet with librarians to discuss the research assignment, the resource requirements, and the research process. This is likely most effective in lower-level courses for students who have not gained research experience (reasons 1–4, 10–12). Many novice researchers struggle with one of two problems: (1) they find too many results for their research topic, or (2) they find few or no resources relevant to their research topic. This lack of information literacy as well as poor research skills cause frustration and increase students’ stress because they spend valuable time searching unsuccessfully for resources (cf. reason 6).
  • Teaching faculty can help students learn better time-management skills by dividing larger assignments into smaller “bites” (reasons 2–3, 10). On the other hand, libraries can help students who are pressed for time by offering extended reference hours during research-critical weeks of a semester. As a supplement or an alternative, librarians should produce short, high-quality, and informative instructional videos, teaching students the basics of research and citations.

None of the points mentioned above are new. So, how does all of this apply to ChatGPT and similar software? ChatGPT is merely a new version of plagiarism that makes it easier for students to plagiarize and harder for faculty to detect. However, the principles of actively preventing such plagiarism, this author believes, remain unchanged: librarians have always known that good information literacy instruction and reference interviews help students succeed ethically.

Even though this author’s primary responsibilities lie in the technical services, he has conducted many in-class instruction sessions and countless reference interviews in which the “lightbulb” went off in the student’s mind. Quality classroom instruction and well-executed reference interviews teach students how to cite correctly (reason 1), help them find high-quality resources that enhance their paper’s quality (reason 2), save them time by leading them to a manageable number of resources that fit their paper’s topic (reason 3), fill them with hope and confidence that they can complete this task by taking the first few steps (reasons 4, 10) and doing it well (reason 11), and teach them how to conduct more research for this and future assignments (reason 12).

Between August 2021 and April 2024, Colorado Christian University librarians conducted 1,995 reference interviews; students provided feedback to 205 (10.28 percent) of these meetings. Of those, one student (0.49 percent) stated that the meeting was “not very helpful,” seven students (3.41 percent) stated that it was “somewhat helpful,” 117 (57.07 percent) said it was “very helpful,” and 80 (39.02 percent) described it as “transformational.” This author believes that approximately 96 percent of students experienced a lightbulb moment when they felt empowered to complete and do the assignment well. For many students, education is merely “a rung in the ladder to success, and not an active process valuable in itself. Because of this, students tend to focus on the results of their research, rather than the skills they learn in doing it” (Šprajc 2017, 33–34). A student who learns how to conduct research well and is empowered to complete an assignment is more likely to see the benefit of education and much less tempted to use AI to complete an assignment.

The Judeo-Christian perspective on plagiarism is clear: do not steal. Generative AI, as the Internet or the smartphone have done, will likely change the world. ChatGPT makes it easier for students to cheat their way through school. This author believes old-fashioned reference interviews and classroom instruction can significantly prevent students from resorting to such unethical practices.

References

Hattingh, Frederick Gerhardus, Albertus A. K. Buitendag, and Manoj Lall. 2020. “Systematic Literature Review to Identify and Rank the Most Common Reasons for Plagiarism.” InSITE 2020: Proceedings of the Informing Science and Information Technology Education Conference: 159–182. https://doi.org/10.28945/4576.

Oxford English Dictionary, rev. 2006. Accessed online, May 19, 2024. “plagiarism (n.).” https://doi.org/10.1093/OED/3549092311.

Paramasivam, Sheela, Tuan Sarifah Aini Syed Ahmad, Irma Ahmad, Nur Fadhlina Zainal Abedin, and Wan Zumusni Wan Mustapha. 2023. “To Copy or not to Copy: Knowledge and Reasons for Plagiarism.” International Journal of Social Science Research 11, no. 2: 105–116. https://doi.org/10.5296/ijssr.v11i2.21217.

Šprajc, Polona, Marko Urh, Janja Jerebic, Dragan Trivan, and Eva Jereb. 2017. “Reasons for Plagiarism in Higher Education.” Organizacija 50, no. 1: 33–45. https://doi.org/10.1515/orga-2017-0002.

Udell, Emily. 2024. “Reading between the Bots: Where librarianship is headed in the age of AI.” American Libraries 55, no. 3–4: 20–27.