Estimated read time: 7 mins

Introduction

Artificial intelligence is here to stay!  Education institutions around the world are discussing what this means for education and assessment. (Colleges seek to 'ChatGPT-proof' assignments | CP24.com) What does this mean for academic integrity in our courses? What are some best practice recommendations for faculty?  

What is AI?

At Georgian College, academic integrity (AI) is outlined in the Academic Regulations (8. Academic integrity < Georgian College Catalogue) based on fundamental values of honesty, trust, fairness, respect, responsibility and courage (ICAI Fundamentals).  

Artificial intelligence (AI) tools (Introduction to Artificial Intelligence (AI) - Academic Integrity and Artificial Intelligence - All Guides at Sheridan Library & Learning Services (libguides.com)), such as ChatGPT, are tools that generate/create outputs based on inputs.  Many are large language models and trained on available material.  They are not search engines or research databases. 

So What?

Almost anyone with access to the internet has the ability to access AI tools. Microsoft and Google are now integrating AI support into their word processing and other software.  AI tools are not going anywhere! We (faculty, staff, students) are exploring when and where it is appropriate to use AI tools and this appropriateness will vary by course, assignment and activity.  Each course/situation is unique and its expectations around AI tools will be unique. 

Now What?

From an academic point of view, here are some best practice recommendations.  Not every suggestion will apply to every course or situation.  Practice and reflection in this new AI world are critical for on-going progress and success. 

Course and assessment design:

• Assess the process as well as the product in evaluations to encourage more time management/less stress near due dates and ideally academic integrity
  • Scaffold assessments into multiple steps with due dates and graded components (Assessment fundamentals - Georgian College)
  • Assign grades for data collection or skill performance in practical/lab activities in addition to any follow-up assignments
• Offer choice of multiple formats (within reason) for assessment to encourage success based on areas of strength, reduced stress about format/process and ideally academic integrity
  • Assess the content of the assessment not the format of the vessel
  • Provide multiple means of action and expression as part of Universal Design for Learning principles (Universal Design for Learning - Georgian College)
• Provide opportunities for students to use feedback and improve on assessment grade to encourage reduced anxiety and stress associated with high-stakes assessment (Formative Assessment - Assessment fundamentals - Georgian College) 

Course communication:

• Lead class discussions about academic integrity early and often and with each assessment.  
  • Don’t assume students know what to expect from a previous semester or a previous assessment.  
  • A simple real-world analogy can be helpful in discussing what is appropriate for your course.
• Explicitly include your expectations around artificial intelligence in your syllabus.  (See resource list)
  • Repeat these expectations in writing and in discussion on each and every assessment.
  • Be clear and transparent with students about the use/avoidance of artificial intelligence.
• Answer (in assessment instructions) students' key questions about each assessment.  Answering these questions helps add value to the assessment and improve their clarity encouraging academic integrity.
  • What do you want the students to do?  
  • Why do you want the students to do the work?  
  • How do you want the students to do the work?  
• Model the behaviour you want to see from students.
  • Support student learning by demonstrating appropriate AI tool use and citation.  APA has recommendation for citing AI work. (Data Sets, Software & Tests - APA (7th ed.) Citing Guide - Library and Academic Success at Georgian College_)
• Have students complete the Academic Integrity Module (AIM) (new release coming Winter 2024). Vice President Academic Home (sharepoint.com) 

Assessment suggestions:

• Require students to include a memo (written or audio) about what they learned and how they did the assessment.
• Include oral conversations/assessments options for follow up/clarification about learning and/or AI use.
• Encourage collaborative and/or independent thinking with In-class assessments/activities.
• To incorporate AI tool use, have students:
  • Reflect on how they used AI and how they edited/modified it for their assessment while including a copy of AI prompt and response as part of assessment submission.  
  • Generate AI prompts and evaluate/correct/annotate AI output.
• To encourage avoidance of AI tools, have students:
  • Include real world examples, personal thoughts/experiences and/or local examples/references in assessment responses as it’s harder for AI tools to effectively include these.
  • Answer hypothetical or scenario-based questions with no simple answer that requires critical thinking.
  • Write potential test questions and answers with references to course material/experiences/activities. 

AI Tool Limitations

As with all technology, AI tools are constantly evolving, and new tools are being released every day.  Current tools, such as ChatGPT, have limitations from an academic perspective.

• AI tools “create” which means sources, citations, and data included in the output may not be real. Each output is unique.
• AI detection tools (such as Turn It In) are not accurate.  
  • It is difficult to prove someone has used an AI tool and it is difficult for an individual to prove they haven’t used an AI tool.  
  • An AI detection tool should not be used as the only evidence of academic misconduct.  
  • Turnitin's Artificial Intelligence Detection Tool will be turned off at Georgian College at the end of 2023.  
• AI tools are not inclusive.  Not everyone can access AI tools and some features have a financial cost. 

Suggested Reading and Resources

• Centre for Teaching and Learning (CTL) Academic integrity - Georgian College  
• Sample Syllabus Statements (variety of institutions)
  • Mohawk College ld.php (mohawkcollege.ca)
  • University of Toronto Syllabus-Language-for-Gen-AI-April-2023.pdf (utoronto.ca)
  • Bryant University Sample Syllabus Statements Regarding AI and Chat GPT – Center for Teaching Excellence (bryant.edu)

Lesson and Assessment Design

• AI in Higher Education Resource Hub | Welcome to TeachOnline
• AI Eroding AI? A New Era for Artificial Intelligence and Academic Integrity | Faculty Focus
• Essential Considerations for Addressing the Possibility of AI-Driven Cheating, Part 2 | Faculty Focus
• How to Talk to Your Students about ChatGPT: A Lesson Plan for High School and College Students | Learning, Teaching and Leadership (wordpress.com)
• A First Response to Assessment and ChatGPT in your Courses | Taylor Institute for Teaching and Learning | University of Calgary (ucalgary.ca)
• Teaching and Learning with Artificial Intelligence Apps | Taylor Institute for Teaching and Learning | University of Calgary (ucalgary.ca) and Teaching-With-AI-Apps (ucalgary.ca)
• Designing assessment for an AI-enabled world | Teaching & Learning - UCL – University College London

Academic Misconduct – Now what?

• Faculty: Managing Unauthorized Use of AI (mohawkcollege.ca) (from Toronto Metropolitan University)
• AI Writing Detection Frequently Asked Questions | Turnitin