Brown DLD Faculty Guides

How Do I Create Assessments That Encourage Original Work and Deter Use of AI?

Updated on

Learning requires effort; without effort, what students learn is superficial and less likely to stick. Yet, humans tend to be reluctant to expend more mental energy than is necessary for a task. These two truths make the role of the instructor one of facilitating and motivating effortful, deep learning so that students can retain and apply what they’ve learned. When students take shortcuts by plagiarizing or with tools like Chatbots, they deprive themselves of the learning experiences that result in sustained knowledge and skills. How then do we motivate students to do original work so that they can reap its benefits? Some answers can be found in tapping into students’ goals and motivations and in designing assessments that mitigate the desire to do unoriginal work. 

Motivating Students To Do Original Work

  • Design relevant, meaningful assignments that connect to students’ goals. When students understand the value of coursework to their goals, they’re more likely to engage deeply with it. Conversely, lack of relevance can influence students' decisions to disengage or withdraw from educational programs (Park & Choi, 2009)
  • Explain the relevance of the assignment to learning outcomes and/or their future career fields. While it's clear to faculty how course assignments build skill and knowledge in the discipline, students might not always make the connection. Create buy-in by explaining how the work will benefit them and advance their goals.
  • Use frequent, low-stakes assessments. High-stakes assessments are a one-time snapshot of students’ knowledge and are less reliable representations of students' overall learning than multiple lower-stakes assessments that occur throughout the course. High-stakes assessments also raise pressure on students, which can motivate cheating. On the other hand, formative assessments like low or no-stakes quizzes help students identify areas they need to revisit, or confirm accurate knowledge, which in turn, promotes learning (Cherem, 2011). 
  • Look for ways to provide more choice in course activities and assessments so that students have a sense of autonomy over their work and learning. Choice can be provided by allowing students to choose a prompt or problem to address in their work or by allowing multiple ways for students to demonstrate their knowledge (.e.g., a poster, paper, or podcast). 

Assignment Design in Times of AI

Advances in artificial intelligence (AI) through tools like ChatGPT, are making instructors reflect on their assignments to see how well they could be handled by AI. While there are advances in technologies that detect AI, these also bring concerns about false accusations since detectors can only offer a probability that text was written by AI. Furthermore, as detection technologies advance, so does the AI, making it a continual arms race. One alternative to detection is to leverage learning principles that motivate original work and draw on best practices in assignment design. These principles often prioritize engaging students in meaningful, process-driven tasks.

  • Scaffold assignments and require a metacognitive component: Have students submit drafts, prototypes, argument maps, or outline of their work and include the rationale for their adjustments or revisions. Ask them to explain how their thinking has shifted as they’ve worked on it. 
  • Design experience-based assessments: Experience is one of the most powerful teachers, and experiential learning allows students to connect course material to real world phenomena and engage in authentic practice of skills. Consider how your course content could be explored through hands-on methods or through observations in non-classroom contexts. During these experiences, students can make observations, take notes, collect data, test hypotheses, apply skills, and so on. Later, they can reflect and discuss how their experiences extended or challenged their understanding of the course content.
  • Use experiential learning activities: Case studies, interviews, simulations, and role-plays are all examples of experiential learning methods that ask students to apply knowledge in authentic, novel situations that cannot be enacted by AI (although AI may provide background information that can assist with them). You can facilitate these hands-on methods during synchronous class sessions or asynchronously using Canvas tools.
  • Use authentic assessments. Authentic assessments mirror how knowledge and skill would be demonstrated in "real life" situations and are often more complex than traditional assignments. For example, in business courses, students might create a marketing or business proposal as a way to demonstrate their learning. In a writing course, students might write a transcript for a podcast that they then record or submit an op-ed to a news outlet. Because of the relevance their goals, interests, or skills they'll use in future contexts, students are likely to be less motivated to rely on AI. Point out to students that these assignments provide enduring evidence of their learning that can be shared elsewhere (e.g. portfolios of student work).
  • Build in a personal connection component to assignments: Weave in personal reflections and connections to course assignments; for example, in an epidemiology course, students might choose to investigate a disease that's affected someone they know, and they can incorporate a reflection on their choice and its significance. Although ChatGPT can feign a personal reflection essay, students will likely be more reluctant to use AI to speak about personal experiences or reflections. Personal connections also bolster retention—we remember things that relate to us and that have emotional significance. Remind students that using their voice is a way to exercise their agency, and that is not something most want to hand over to AI!

Example Assignment that Motivates Original Work

In Irene Glasser’s Anthropology of Addiction course, students investigate the role of culture in drinking habits and norms. As part of their coursework, students engage in experiential learning by interviewing someone from a different culture about drinking in their native country, and by attending local AA meetings. Students write up their notes from the interview and engage in data analysis to identify themes.

Image of an assignment prompt that asks students to interview someone about drinking habits in another culture
Previous Article Proctoring Software: Policies & Practices
Next Article Overview: Learning Technologies and Academic Integrity
Still need help? Contact DLD