Generative Artificial Intelligence Guide for Faculty

The opportunities and challenges associated with GAI are not much different from those faced by educators in the face of any other technology revolution. In recent times, educators have had to adapt to Google, Wikipedia, calculators, laptops and cell phones making their way into the classroom, bringing with them the potential for cheating. During the Covid-19 pandemic starting in 2020, many educators had to adapt to teaching fully online for the first time in their careers. This page offers general advice about critical assignment design that can help reduce reliance on GAI.

Page Contents:

• Project Recommendations
• Assignment Methodology
• Sample Assignments
• Evaluating Assignments to Account for GAI Use

Project Recommendations
These tips can help you write assignments that promote critical thinking, engage creativity, help students add their voice to their discipline, and hold them accountable for what they submit. By implementing these solutions, faculty members can encourage students to use generative AI tools as a tool for learning while still ensuring academic integrity and preventing plagiarism.

• Assign open-ended or personalized questions. Encourage students to use generative AI tools for idea generation rather than direct answers. Ask them to apply personal experiences, critical thinking, or analysis, making it difficult for AI-generated responses to serve as substitutes.
• Incorporate problem-based learning. Design assignments that require students to solve real-world problems by applying knowledge rather than relying on AI-generated information. Consider multipart assignments that increase in complexity. 
• Assign group projects. Assign group projects and assignments that require students to collaborate and share their ideas and knowledge, ensuring discussion and deeper engagement rather than reliance on AI.
• Incorporate critical thinking exercises and skills. Have students assess AI-generated content for accuracy, bias, and reliability, cross-checking facts and expanding upon AI-generated insights with additional research.
• Provide clear ethical guidelines. Educate students on proper and improper use of AI tools, reinforcing that AI should enhance learning, not serve as a shortcut to completing assignments.
• Use staged or incremental assignments with accountability. Break large assignments into milestones or progress submissions that are discussed in real time. Require students to defend their sources, articulate reasoning, and reflect on how AI tools were used at each stage.

Assignment Methodology

There are many good methods and theories about drafting prompts for academic projects. Ideally, a well-constructed project should help connect students to their intended profession. A well crafted assignment will foster critical thinking and give students an intriguing purpose to their work that can also reduce opportunities for any type of cheating.

One method that is easy to emulate is the is RAFT Method of assignment construction as articulated in the book Engaging Ideas by John C. Bea and Dan Melzer. The RAFT model encourages task-based assignment writing that requires students to think and write in a way that models the processes and outputs of professionals in their discipline

Source: Bean and Melzer, Engaging Ideas, 3rd ed.

Bean and Melzer illustrate variations in assignment design with an example of the different ways in which a prompt could be written and suggest that we consider:

• What differences in thinking are apt to be encouraged by each option?
• What are advantages and disadvantages of each option?

With regard to GAI, we might further consider:

• Which options have elements that you could allow students to enhance with GAI?
• Which options are most likely to discourage cheating?
• What should happen before any of these assignment variations is presented to students to ensure that they have the skills and confidence to complete it ethically and competently?

Source: Bean, Engaging Ideas, 2nd ed., pp. 92-93. Click image to enlarge.

Sample Assignments

Here are samples of assignment types and milestones that either engage with GAI or make it difficult for GAI to generate successfully. 

• Have Students Assess AI-Generated Output: Have students examine AI output for accuracy, bias, and reliability, cross-checking facts and expanding upon AI-generated insights with additional research.
• Journaling: Student provide accounts of how they approached each stage of a project. Students should report in chronological order how they did the work, where they did it, what results they got at each stage, how they adjusted, whom they asked for advice, how they incorporated that advice, etc.
• Oral Presentations: Students respond live to questions not provided in advance.
• Observational Studies: Students produce lab reports, field notes, or creative works.
• Primary Source or Data Analysis: Students use historical documents, data sets, or other primary source material to contextualize a concept.
• Annotate Sources: Have students find subscription-based material held by the library (firewalled) and answer questions that AI will not be able to generate. 

Evaluating Assignments to Account for GAI Use
Rubrics can be structured to account for originality of work. The examples below provide elements of rubrics that can be incorporated into assignment design.