How can I support students in learning the basic skills of effective generative AI use?
The AI Toolkit is a great resource to share with students that includes a wealth of information about proper and effective use of generative AI, but if you are looking for something more interactive and embedded directly in your Brightspace course, we have that too. The DePeters Family Center has created a Generative AI Basics module that any faculty can copy directly into their own Brightspace course. Email cite@sjf.edu to request access.
How can AI support my teaching?
As the academic landscape evolves, so too does the role of technology in enhancing teaching methodologies. In this section, we uncover the key question of "How can AI support my teaching?"
Below you'll find specific uses of AI tools that can support faculty's teaching. Below that, you'll see additional resources.
Imagine you've been given a set of outcomes or standards that you will need to integrate into a course that you will teach next semester. You've never taught this course and you're unsure if the textbook and old syllabi that you were given truly speak to how you would like to teach the course. Perhaps you're wondering: Where do I start when designing this course?
AI tools can take outcomes or standards and "unpack" them so that key concepts are easily identified. These concepts can then be used to generate essential questions (which can provide a structure your course) and craft learning objectives. Please see the "Generating Essential Questions" and "Crafting Learning Objectives" sections below for more information.
Example Prompts for Unpacking Outcomes:
Can you conceptually unpack the following outcome: [Paste outcome here].
Depending on the response you may want to ask the following questions:
Can you please expand upon what you mean by [concept]?
What are some additional questions that may be useful in conceptually unpacking this outcome?
What are some specific examples of activities that students can show their understanding of [concept]?
Please see the "Generating Essential Questions" section below for more ways in which outcomes can be further unpacked.
You've asked the AI tool to unpack your outcomes or standards and you're now wondering where to go next. An important piece of high-quality curriculum design is that of essential questions. These are questions that reflect the most historically important issues, problems, and debates in a field of study. They are meant to provide structure for a course by framing each week of the course with key inquiries that are inherent to the discipline.
Example Prompts for Generating Essential Questions:
What are some essential questions that relate to the outcome's key concepts?
If you didn't start with unpacking outcomes, you can ask the system the following question: Can you please generate some essential questions that are necessary to teach students in a college-level course about [subject area]?
Now that you've conceptually unpacked the outcomes or standards and you've generated some essential questions, you're now ready to ask the system to craft some learning objectives. It's important to outline that outcomes are the desired results of the learning process, while objectives are the "how" to achieve the desired results. When asking an AI tool to craft learning objectives, they will most likely give you skilled-based objectives that include performance verbs, such as explain, define, identify, analyze, evaluate, construct, etc. These are particularly helpful when deciding what you would like students to be able to do and be skilled at by the end of the semester.
Word of Caution: AI tools can provide a fantastic starting point for crafting learning objectives. However, it's best to review each of the objectives that the system provides to ensure accuracy and alignment with the subject at-hand. As the content expert, you'll be able to determine if the objectives are worth using and if follow-up questions should be asked of the system to offer clarity.
Example Prompts for Crafting Learning Objectives:
If you have followed the sequence of asking the tool to unpack outcomes and generate essential questions, you can ask the following question: Can you provide some learning objectives that are aligned with these concepts and essential questions?
If you didn't start with unpacking outcomes, you can ask the system the following question: Can you please generate some learning objectives that are necessary to teach students in a college-level course about [subject area]?
Generative AI tools can provide important insights and suggestions to help instructors identify suitable instructional methods to enhance engagement and learning in their face-to-face, hybrid, or online classes. It is essential, however, for instructors to adapt and tailor these methods to suit the specific context and needs of their students and subject matter. Additionally, instructors should draw on their own expertise and experience to design an effective and inclusive learning environment. Generative AI tools can be a great starting point when trying to identify high-quality instructional methods.
Word of Caution: Generative AI tools can provide a fantastic starting point for identifying potential instructional methods. However, it's best to review each of the methods that the system provides to ensure alignment with the learning objectives. As the content expert, you'll need to take some time to determine if the instructional methods are an effective way in which to increase student knowledge and understanding of your disciplines key concepts and skills.
Example Prompts for Identifying Instructional Methods:
What are some effective instructional methods for teaching [subject/topic] in a college-level class?
Can you suggest interactive teaching strategies suitable for a college course on [subject]?
Please provide innovative instructional techniques to engage students in a college-level class on [subject].
What are some active learning activities that can be incorporated into a college course on [topic]?
Can you recommend technology-based instructional methods suitable for teaching [subject] at the college level?
How can I incorporate group discussions and collaborative learning in my college class on [topic]?
What are some hands-on experiential learning opportunities for a college course focused on [subject]?
Please suggest methods for promoting critical thinking and problem-solving skills in a college-level class.
How can I use case studies and real-world examples to enhance learning in my college course on [topic]?
Can you propose assessment strategies that align with active learning methods in a college-level class?
By using these prompts, generative AI tools can provide insights and suggestions to help instructors identify appropriate formative and summative assessment opportunities that align with their learning objectives and teaching approaches. Pre-assessments take place before the learning process to determine students' prior knowledge (e.g., surveys, short quizzes, entrance tickets, etc.); formative assessments take place during the learning process (e.g., asking students questions, classroom discussions, short quizzes, polls, etc.); summative assessments take place at the end of the learning process (e.g., traditional tests, research papers, portfolios, presentations, etc.). It is important that instructors adapt these assessments to suit the specific needs and learning outcomes of their students and subject matter. Additionally, instructors should consider the validity, reliability, and fairness of the assessment methods to ensure accurate evaluation of student progress and achievement.
Word of Caution: Similar to identifying instructional methods, generative AI tools can provide a useful starting point for identifying potential assessment methods. However, it's best to review each of the methods that the system provides to ensure alignment with the learning objectives. As the content expert, you'll need to take some time to determine if the assessment methods are an effective way in which to measure student knowledge and understanding of your disciplines key concepts and skills.
Example Prompts for Identifying Assessment Opportunities:
What are some pre-assessment techniques suitable for gauging students' prior knowledge in a college-level course on [subject]?
What are some formative assessment techniques suitable for gauging student understanding in a college-level course on [subject]?
Please suggest creative formative assessment strategies to promote student engagement in my college class.
Can you recommend formative assessment activities that can be easily integrated into a college course on [topic]?
What are some formative assessment tools or technologies that can provide timely feedback to students?
How can I design effective formative assessments to monitor student progress throughout the semester in my college class?
Can you propose summative assessment methods appropriate for evaluating student learning outcomes in a college-level course on [subject]?
Please suggest summative assessment strategies that align with active learning approaches in my college class.
What are some authentic summative assessment tasks that can assess students' practical skills and knowledge in a college course on [topic]?
How can I use rubrics effectively for summative assessment in a college-level class? (See more about constructing rubrics below)
Can you recommend ways to balance formative and summative assessments to ensure comprehensive evaluation of student learning in my college course?
You've noticed that students enjoy your end-of-semester project, but sometimes express that they feel lost at times. They want to know more about what you're expecting so that they can do as well as possible and earn a high grade. This is where rubrics can remove the mystery for students and provide more credibility to faculty's grading decisions. AI tools can help you construct rubrics in a flash. Feeding these tools specific prompts can result in four-column rubrics that consist of an array of criteria that will support both you and your students.
Word of Caution: AI tools can save you a significant amount of time typically needed to design and develop rubrics. It's highly suggested, however, that you take extra time to review the generated rubric and consider how it aligns with your assessment (i.e., a research paper, presentation, project, etc.). It's important to review and add further details to the rubric to ensure that the content is accurate, that it's clear and practical, and that the criteria adequately measures the goal being assessed.
Example Prompts for Constructing Rubrics:
Please create a four-column rubric for this assignment: [paste in assignment description]
For example, say you prompted the generative AI tool to write a case study that focuses on three big concepts that students need to know and understand in your course. Then, you asked the system to write some rough draft directions and an assignment description that students can follow to complete the case study assignment. It then may be worthwhile to ask the system the prompt above and paste in the assignment directions and description to create a four-column rubric. This will most likely result in a rubric that is much more aligned with the assignment goals.
For a more generalized rubric, you can ask the following: Can you create a four-column rubric that is associated with a [assignment type, e.g., research paper, presentation, etc.] where students need to be skilled at [list skills, e.g., defining, explaining, analyzing, communicating, evaluating, etc.]?
As a faculty teaching in the age of generative AI, one of the hardest questions to answer is, ‘what do I do if I suspect my student used generative AI inappropriately for their assignment?’
St. John Fisher University does not encourage the use of generative AI detectors to identify student misuse of AI tools. The generative AI detector included in our Turnitin platform has been turned off effective August 28, 2024. This decision was made based on growing evidence on the lack of accuracy, including both false-positives and false-negatives, in the reports, bias in the reports towards non-native English writers, and the ability for students to easily avoid detection with these tools creating a false sense of security for instructors.
In this section, we aim to provide guidance for faculty on what to consider to avoid student misuse of generative AI tools and what to do if you suspect generative AI may have been used inappropriately. Please consider the use of many or all of these strategies. No one strategy in isolation will likely be useful in these situations.
This should be clearly outlined in the syllabus and communicated in a follow-up discussion either in-class or virtually. You should also offer a chance for students to ask clarifying questions. Generative AI features are prevalent in many tools, so students may be using these features without fully knowing and/or may need to clarify what and why they are using a particular tool.
You may compare prior work to current submissions, as well as in-class writing compared to other submissions. Identifying these differences in style is also a great starting point for a discussion with the student. However, changes in writing style alone may not indicate any misuse of generative AI. Changes in writing style may be necessary depending on the assignment and instructions given.
Generative AI tools are not intelligent, so while writing may sound accurate, the logical flow or evidence for an argument may not be present or may be unclear.
Generative AI tools will attempt to create citations but are often either partially or totally inaccurate. Check the DOI links or other components of the full citation, like author, journal title, etc. for accuracy. If inaccuracies are found, talk with the student about their sources, where they found them, and to bring copies of the original works with them to talk with you.
You will not get the same result twice, and there are several tools to try this with, but you may see patterns in the results you receive that also appear in student work. Reading a lot of AI generated text will help you identify it more quickly, as it will often follow similar patterns and styles in its output.
Assume the best of students, not the worst. Students will likely explain if they used a generative AI tool and how. This can be a great opportunity for discussion on what are appropriate and inappropriate uses of generative AI tools in your course. As it is common not to accuse a student of plagiarism in a first-year writing course when they are learning new writing and citation skills, we should treat all students in a similar way since generative AI and its use as part of the learning process is still very new for most faculty and students. Use the conversation with the student as a teachable moment to support the student’s evolution with building AI skills in an ethical and safe manner.
The Academic Integrity Committee is a place to go for individual consultation on a specific situation and can provide additional context-specific guidance.
Instead of trying to detect generative AI use, we suggest transitioning your assessments to activities that are naturally more AI resilient, meaning even with access to generative AI tools, they will be useful measures of student learning and performance. Explore the options below for ideas:
As generative AI becomes more accessible and powerful, it also poses new challenges and dilemmas. While generative AI can be a helpful tool for learning and research, it can also be misused to cheat, plagiarize, or generate low-quality work. Therefore, it is important to have an open and honest conversation with your students about the ethical implications of using AI for their work.
Some questions you can ask your students are:
- What are some examples of appropriate and inappropriate uses of generative AI for your assignments?
- How do you decide whether to use a generative AI tool or not for your work?
- What are some potential benefits and risks of using generative AI for your work?
- How does using generative AI for your work affect your learning outcomes and academic development?
- How does using generative AI for your work affect the value and credibility of your work?
- How does using generative AI for your work affect the rights and responsibilities of other students, instructors, and researchers?
- How do you cite and acknowledge the sources and tools you use for your work, including generative AI?
- What are some ethical principles or guidelines that can help you use generative AI responsibly and respectfully for your work?
The goal of this discussion is to encourage critical reflection and informed decision-making. You can use this discussion as an opportunity to clarify your expectations and policies regarding generative AI use, as well as to provide feedback and support to your students. By engaging in this discussion, you can help your students develop ethical awareness and digital literacy skills that are essential for academic success and lifelong learning.
One way to create assignments that emphasize evaluating the accuracy of AI generated information is to ask students to compare and contrast different sources of information on the same topic, such as news articles, academic papers, Wikipedia entries, or AI generated summaries. You can ask students to identify the strengths and weaknesses of each source, such as the level of detail, the quality of evidence, the potential bias, or the reliability of the author. You can also ask students to explain how they would verify or cross-check the information from each source, such as by using other sources, checking the references, or using online tools. By doing this, you can help students develop critical thinking skills and learn how to assess the credibility and validity of AI generated information.
Another way to create assignments that emphasize producing original work and avoiding generative AI shortcuts is to incorporate real-world projects that connect students with authentic audiences and contexts. For example, you can ask students to design a website, create a podcast, produce a video, or write a blog post for a specific purpose and audience. You can also ask students to collaborate with peers, experts, or community partners in these projects, and to reflect on their learning process and outcomes. By doing this, you can help students develop relevant skills and knowledge that are not easily replicated by generative AI tools and motivate them to engage more deeply with the content and the task.
You can also use in-class activities that require students to demonstrate their understanding and skills in a more interactive and dynamic way to showcase their skillset. For example, you can design quizzes, games, discussions, simulations, or experiments that challenge students to apply their learning to new situations, solve problems, or generate questions. These activities can help you monitor students' progress, provide feedback, and adjust your instruction accordingly. They can also make it harder for students to rely on generative AI tools, as they often involve spontaneous responses, collaborative work, or multiple modes of expression.
Oral exams and presentations are another way to assess students' learning without relying on written assignments that can be easily done by generative AI. Oral exams and presentations require students to articulate their knowledge, understanding, and arguments in a clear and coherent way, using their own words and voice. They also allow the teacher to ask follow-up questions, probe deeper into students' reasoning, and check for comprehension. Oral exams and presentations can test students' critical thinking, communication, and presentation skills, as well as their ability to synthesize and apply information from various sources. They can also be more engaging and motivating for students who prefer oral expression over written expression.
Scaffolded writing is a strategy that breaks down a complex writing task into smaller and more manageable steps, providing guidance and feedback along the way. Scaffolded writing can help avoid students using generative AI to do their work by making the writing process more transparent and personalized, and by requiring students to demonstrate their own thinking and creativity at each stage. For example, a scaffolded writing assignment might involve brainstorming, outlining, drafting, revising, and editing, with specific instructions and criteria for each step. The teacher can check students' work at each step, provide feedback, and monitor their progress. This way, students are less likely to rely on AI to generate or edit their work, since they must show how they developed their ideas, arguments, and evidence from scratch. Scaffolded writing can also help students improve their writing skills and confidence, as they learn to plan, organize, and refine their work systematically and effectively.
Personalized assignments allow students to choose a topic, genre, or format that suits their interests, motivations, and goals. This way, students are more likely to engage with the assignment and produce original and authentic work, rather than relying on AI to generate or edit generic content. Personalized assignments also challenge students to use their own voice, perspective, and creativity, which are difficult to imitate by generative AI tools. Additionally, personalized assignments make it easier for teachers to detect plagiarism or AI-generated content, since they can compare students' work with their previous submissions and their known preferences and abilities.
Peer review involves students giving and receiving feedback on each other's work, which can help them improve their writing skills, learn from different perspectives, and develop critical thinking. Peer review also requires students to justify their choices and arguments, which can expose any inconsistencies or errors in AI-generated content. Collaborative work involves students working together on a common project or task, which can foster communication, cooperation, and creativity. Collaborative work also encourages students to share their ideas and opinions, which can reveal their personal style and voice, and make it harder for them to use AI-generated content without being noticed.
Teaching with AI Resources
This a guide for educators who use ChatGPT in their classroom. In particular, this guide includes suggested prompts, an explanation of how ChatGPT works and its limitations, the efficacy of AI detectors, and bias.
How can educators respond to students presenting AI-generated content as their own?
This article was written for OpenAI's "Educator FAQ" section that uncovers ways in which teachers can use ChatGPT safely in the classroom. When it comes to matters regarding AI-generated content and students presenting it as their own work, this article is well worth the read.
Readings
Teaching in the Age of AI - Center for Teaching, Vanderbilt University
This resource explores what AI is and where it can be found. Additionally, it describes ways in which educators can harness AI tools in their teaching and improve student learning. It also uncovers how to craft assignments that deter unauthorized AI use and explains how academic integrity relates to AI tools as well. Lastly, it describes what specific resources are available for instructors who want to engage with AI tools.
Maximizing your Course Success: Utilizing ChatGPT & Prompt Engineering
This e-book aims to help readers learn how to communicate effectively with ChatGPT to maximize the success of their online courses. It covers various aspects, including prompting tips and hacks, the prompting framework designed for course creators, and expertly crafted prompts for all stages of creating an online course. By utilizing the information in this e-book, readers will gain the knowledge and techniques to unleash ChatGPT's full potential and elevate the quality of their online courses to new heights. For additional information about writing quality prompts, please see the "How to Write Good AI Prompts" page.
Podcasts & Videos
The Tea for Teaching podcast series, hosted by John Kane, an economist, and Rebecca Mushtare, a graphic designer, revolves around informal discussions focusing on innovative and effective practices in teaching and learning. Both hosts are affiliated with the Center for Excellence in Learning and Teaching at the State University of New York at Oswego. Throughout the podcast, they explore various topics related to education (including AI), sharing valuable insights and ideas to benefit educators and learners alike.
The Teaching in Higher Ed Podcast is hosted by Bonni Stachowiak and airs on a weekly basis. It explores various topics pertaining to teaching in higher education, encompassing excellence in teaching, instructional design, open education, diversity and inclusion, productivity, creative teaching methods, educational technology, and blended learning. There are a number of episodes strictly dedicated to uncovering how AI is influencing education and how to harness it.
How to Use AI in Research: For Students
Lavery Library’s Research Guide, AI Tools and Resources, includes a section for students. Topics include how to get help from Lavery Library, citing AI, evaluating sources suggested by AI-driven chatbots, how to track down sources suggested by AI tools, how to use AI to find scholarly sources, and more.
How Professors Can Check for AI Cheating in a Sensible Way
This video created by George Fox University provides practical tips for faculty to consider using to deter students from using generative AI in inappropriate ways.