Summary

As AI tools become more prevalent and accessible, it is important to guide users toward responsible and ethical use. This lesson will ask students to use generative AI to think through topic development and research planning. The lesson will introduce AI platforms and ask students to prompt two separate AI platforms to ask them questions about their research topics. Students will evaluate and compare responses, while also adapting and adjusting their own topics based on those responses.

Author

Beth Carpenter, University at Buffalo

Learning Outcomes

• Students will be able to perform a basic request from a generative AI tool

• Students will be able to evaluate responses generated

• Students will be able to examine and edit their research questions and topic ideas   

Audience

First-year undergraduates, transfer students

Curricular Context 

This 50-minute lesson will ideally take place in an early college/university course. Many institutions have a required first-year course, and this session would fit well into either a one-shot session or in scaffolded instruction for that course. Independent of a required class for all students, this session would also work well with any class that has a research project as part of the syllabus. The lesson itself is independent of previous library or information literacy instruction, but will reinforce concepts that are taught in both. This lesson would be best taught in a synchronous format, and could work well both in-person or virtual, but all students should have access to a computer or device (such as a tablet). Paramount to this lesson’s success will be buy-in from the course instructor, as the use of generative AI is still a hot-button issue. 

Preparation

 Equipment Necessary:

• Instructor computer and/or projector (if in-person)

• Laptops or devices for all students

• Access to at least three (3) generative AI platforms - at least one should not require account creation or login. Listed below are some examples:

  • ChatGPT

  • Perplexity (no log-in information required)

  • Gemini

  • Cohere Coral

  • Claude

  • Microsoft Copilot

For students who do not want to make their own accounts, Perplexity will work for the lesson. ChatGPT will also provide answers without needing an account to log in, but there may be a limit to the number of answers it provides before asking for a login.

• Instructors may want to share the links to the class ahead of time to allow students to set up accounts beforehand. This will save time during the introduction section of the lesson 

• If working with an instructor, embedding either a link to the exit ticket, or embed the exit ticket, into the course’s LMS

• Nearpod or other interactive teaching platform (not required, but helpful)

• Dry erase boards/Padlet - shared writing space

If accessibility tools are needed or desired, here are some suggestions:

• Screen readers: JAWS, NVDA, or VoiceOver

• Speech-to-Text Software: Dragon NaturallySpeaking, or built-in systems such as Windows Speech Recognition or Apple Dictation

• Voice Assistants: Siri, Google Assistant, Alexa

• Magnification Tools: ZoomText, or built-in magnifier

Instructors may want a few example prompts in mind - both good and bad examples - to demonstrate to students. Questions can, and perhaps should, be tailored to the class visit, but if it’s a general instruction session, having some ready-to-go ideas can be helpful.  My go-to's have been:

• Opening prompt: 

  • Ask the generative AI platform of your choosing to write a haiku on any subject

  • Ask the generative AI platform of your choosing to write a short sentence on any subject

• Main Exercise prompt:

  • Can you ask me questions about romantic comedies? / Can you ask me research questions about romantic comedies

  • I’m writing a research paper on philosophy, can you ask me questions about philosophy?

  • Can you ask me research questions about climate change

Supplementary Materials 

Lesson Outline

Introduction (5 minutes)

The instructor will welcome students to the class, and prompt them to log onto computers/turn on computers/access devices.

If there are time concerns with having enough time to sign up for AI tools during the class period, the instructor should share links ahead of time, and then prompt students to access the generative AI platforms they have chosen.

If the instructor has not shared links ahead of time, write or share links via projector or whiteboard. If using Nearpod, Mentimeter, etc, this can also be a time to gauge student familiarity with AI using a poll, or other participatory feature.

The instructor should also hand out the worksheet, or share a QR code or link to the worksheet/exit ticket. This could be a paper worksheet, it could be a quiz/survey integrated into your institution’s learning management system, or a Google or Microsoft form shared via QR code or easy-to-share URL (TinyURL, bit.ly, etc).

Warm-Up Activity (5-7 minutes)

Ask students to prompt the AI tool of their choice to write a short haiku/short sentence on any topic. Have a few students share out, and discuss how easy or difficult it was to navigate the platform. If they had time to use multiple platforms, there can be discussion about the different responses, and if that tells the user anything about the systems.

Demonstrate Using AI Tools for Topic Development (5-7 minutes)

Explain the type of prompting that will occur in class - asking the AI tool to ask questions, rather than just give information. This process is mirrored when students begin research, both in Google or library resources. Basic searches yield some results, but refining keywords, and learning vocabulary, will help focus those searches.

The instructor will choose two AI tools to demonstrate using them for topic development. It can help to start with a broad prompt (e.g. “Please ask me questions about romantic comedies”) and then adjust the question (e.g. “Please ask me research questions about romantic comedies”) to show how little adjustments can make a difference to the answers. Do the same with the second platform, and point out any differences you see; ask the students if they notice any differences or similarities to the approach.

Students Use AI Tools On Their Own (10 minutes)

When the demonstration is complete, let the students begin their own work using the AI tools of their choosing. If using a tool such as Nearpod, students can enter their thoughts in a Collaborate Board. If not, then have students make notes on dry-erase boards or on a padlet with their topics. Being able to have a shared writing space to refer to for discussion will help students share out.

Have students also keep track of their prompts on the worksheet, along with any adjustments they made.

Demonstrate refining the prompts to the tools (5 minutes)

Narrowing down to one AI tool, demonstrate adjusting the prompt using previous responses. In the romantic comedy example, if “When Harry Met Sally” or “Nora Ephron” gets mentioned, adjusting the prompt to ask for research questions regarding those specific elements; the same might work for a specific trope or commonality. The instructor can feel free to prepare this adjustment ahead of time, or go on the fly, but note that answers given by AI will change.

Students Work On Their Own (10 minutes)

Give students more time to work and adjust prompts. Have them make note of new concepts or terms they’ve learned, and if there are any changes to their research topic since starting the exercise.

Discussion (5-7 minutes)

Pulling from the Nearpod/Padlet/dry-erase boards, ask students to share their thoughts - did they have a favorite tool from the ones they used? Did they feel successful with the responses they got? Did they learn anything new about their topic?

Conclusion (2-3 minutes)

End by reminding students about the worksheet/exit ticket. If digital, ensure students submit before leaving.

Assessment

If using an interactive teaching platform, like Nearpod, reports can be pulled from the session to gain qualitative and anecdotal information from the students. The exit tickets can also act as a way to gain further data from the lesson for iteration and development.

Adaptability 

This lesson would work well for students new to doing research, and would work with many disciplines. It could be adjusted for a higher-level course by simply changing the prompts to be more specific to the discipline. A curricular context that may need more adaptation would be when working with international students. AI tools may be extremely helpful if English is not their first language, but may also cause confusion around this process.

This could also be taught in conjunction with an introduction to searching, if the class period is longer, or if there’s a workshop to be held. This would show how the skills learned using the AI tools adapt to the research process, aligning with the ACRL Frame “Searching as Strategic Exploration.”

Reflection

The greatest difficulty in developing this lesson was wanting more time - there is so much to discuss with AI, such as ethical use, privacy, etc, and there can be mentions of it, but it’s difficult to devote as much time as I’d like to all the areas that are worth talking about. 

In addition, the lack of AI platforms that are available for use without account creation is an ongoing struggle. Many students and faculty do not want to create accounts for things where their information is shared with the platform, and what their information will be used for is unclear.