Re-Introducing LIB100

The Z Smith Reynolds Library at Wake Forest University has been offering iterations of the credit-bearing library research introductory class LIB 100 for over two decades. For the 2025-26 academic year, we filled 22 sections of LIB100 courses with total enrollment around 317 students. This number does not include our other LIB courses or First Year Seminars, which would equal 33 sections of LIB courses with total enrollment around 485 students. This is typical for our course offerings and enrollment. Our waitlists are regularly entirely full, especially for the online, Spring, and second half sections.

Our most popular course is LIB 100: Academic Research and Information Issues, which introduces students to research in academic libraries as well as online. As both our library and the class continue to evolve, we wanted to reintroduce you to the LIB 100 of this particular juncture. While the centrality of the library remains a core organizing principle of the class, it’s probably not the LIB 100 you remember! Over the past three years, we have made numerous revisions to this course to incorporate more AI literacy, more news literacy, and more social media literacy concepts (think AI slop and misinformation, recognizing content about news vs, actual news, and influencer culture online). We feel these revisions acknowledge the challenges we’ve noticed over the past few years with high levels of student skepticism about information, as well as social media content and AI outputs. We hope that our course will provide students with more confidence in their ability to locate, verify, evaluate, and understand information and the many forms it takes, and ultimately set students up for success in nearly every upper-division course, in graduate school, and in any career or life choices that ask them to make decisions based on evidence.

AI Literacy in LIB100

One of the biggest areas of growth for LIB100 in recent years has been in the area of AI Literacy. AI Literacy itself is an evolving concept, and while early frameworks for AI literacy worked to create sets of general competences, some recent scholarship has argued that, “AI literacy cannot be reduced to a universal framework” and instead should be viewed as a discipline-specific, situated literacy practice. (And really, it’s probably both). Regardless of where the AI Literacy debate eventually shakes out, our situated literacy practice, information literacy, has a lot of overlap with many “general” AI Literacy competencies, knowledge practices, and dispositions. This means we have spent the last three and a half years iteratively designing and redesigning LIB100 to incorporate more AI Literacy as generative AI has grown in capability and influence.

In the early days of generative AI, our first foray in AI literacy was a simple source verification assignment – is this ChatGPT generated source a hallucination? If so, why did that happen and how do we find real sources instead? Now, over three years later, AI Literacy concepts are present in over a third of our class sessions. We will loosely divide them into three main categories, which we will spread over three separate blog posts:

• Getting Started: Brainstorming, Research Questions and Prompting, and the Scholarly Conversation
• Finding Information: Source Discovery, Verification, and Access
• Evaluating Information: Evaluating AI Output, Summaries, and Synthesis

We also take care to incorporate Critical AI Literacy concepts into conversations where it makes sense, as we will discuss some of those ideas in this series. We will also take a closer look at some “conventional wisdom” around using AI for research and whether or not those ideas are “holding up” as time passes, generative AI improves, and more research is conducted around AI Literacy.

Getting Started: Research Questions, Prompting, and the Scholarly Conversations

AI for Brainstorming: Is that Really a Good Idea?

One of the most important components of AI Literacy is actually knowing when it’s a good idea to use AI and when it’s probably not a good idea, which is why we think it’s so valuable to start off LIB100 with a conversation about using AI for brainstorming. In LIB100, we’ve long taught that successful research projects start with a good research question. But developing one is harder than it sounds—students consistently report that it is the most difficult part of the research process Head, 2010). Because they often lack the subject expertise to know where lines of inquiry and knowledge gaps already exist, narrowing a topic can feel overwhelming and anxiety-provoking. Many students worry about choosing a “bad” topic and then finding “nothing.” So, this seems like a good place for AI to help us out…. right?

Conventional wisdom around AI has recommended that students use generative AI for topic brainstorming. During a recent session at LOEX, Lane Wilkinson pointed to dozens of library Libguides giving this exact advice (side note: ours doesn’t). Recently, some educators like Wilkinson have begun to push back on that recommendation, essentially arguing that using generative AI for brainstorming undermines students’ epistemic autonomy and leads to homogeneity and the erosion of creativity

While the appropriate role of AI in brainstorming is still up for debate, in LIB100, we’ve generally discouraged using AI at the brainstorming stage because we still see value in students doing the messy work of developing their own topics into productive lines of inquiry. After all, how can we evaluate whether AI has generated a good research question if we have never learned how to develop one ourselves? Instead, students refine their topics into questions through a series of in-class activities. Only after students and instructors have settled on a question do we invite them to see what generative AI produces and evaluate how well AI did.

Asking Better Research Questions in the AI Era

When discussing the characteristics of a strong research question, one of the central principles is avoiding leading questions. This has become even more important in the AI era, where prompting bias can lead generative AI tools to accept assumptions embedded in a prompt and then selectively emphasize evidence that appears to support those assumptions. For example, when asking ChatGPT for evidence that supports opposing sides of the “peptides” debate, we found that ChatGPT would give the same (real) scholarly source, but summarized the source differently based on the way we framed our prompt. Essentially, the wording of a research question can influence not only what evidence an AI tool presents, but also how that evidence is interpreted. For LIB100, this means we have been placing greater emphasis on developing “neutral,” open-ended research questions, recognizing the influence of prompt framing, and verifying AI-generated summaries against the original sources.

Prompt Engineering for Undergraduate Research

Even with instruction, students often continue to struggle with traditional database search techniques, like keywords and Boolean operators. One benefit of AI tools is that they allow students to search using natural language. While early recommendations focused on teaching students relatively complex prompting methods, some recent studies have noted that too much context (aka “prompt bloat”) can lead to worse performance from AI tools. It may be that the level of detail needed for prompting is context-dependent (e.g. what tool, what question, etc…). In our experience, it tends to be that in many cases, students achieve better, faster results with more concise prompts. For LIB100, this has meant spending our instruction time on helping students clearly articulate their information needs, evaluating the AI output, and then use follow-up prompts to further refine their results as needed. In fact, in some cases, the follow-up prompts deserve as much attention as the initial prompt itself.

Scholarship as Conversation

Finally, one of the central early learning outcomes of LIB100, and of information literacy more broadly, is that scholarship is an ongoing conversation. As new research emerges, our understanding of topics evolves, expands, and sometimes changes. Researchers studying the same issue may reach different conclusions or approach it from different perspectives. In LIB100, our goal is to locate and engage with sources that reflect the broad strokes of the conversation. This means identifying common themes, areas of consensus, and significant points of disagreement, while recognizing the difference between well-supported viewpoints and fringe claims.

The research surrounding AI has provided a particularly useful illustration of this concept. Early on in the course, we examine an MIT study published last year that claimed students using AI to write papers were accumulating a form of “cognitive debt.” To illustrate the ongoing conversation around the topic, students read David Brooks’ New York Times opinion piece on the study, which asked whether AI will make us “dumber,” and other researchers’ response to the study which pointed out limitations in the study’s design and sample size. In LIB100, we use this debate to help students see how scholarship actually works in practice – research findings are published, scrutinized, challenged, replicated (or not), and refined by other researchers. Students also generally find this debate relevant and come with a wide variety of opinions, which makes class more fun and engaging for everyone involved.

We’ll wrap this up, but please stay tuned for part two of this three part series in a few weeks where we will chat about source discovery, verification, and access!