Designing your assessments and exams to make them Generative AI compatible or resilient is part of a process that takes time and might involve several iterations.
Adopting an iterative approach
Each stage of the assessment process introduced below can be examined by the following questions (from Digital Education Council, The Next Era of Assessment):
- What is now possible because of GenAI?
- What approach should I adopt in reaction to GenAI?
These questions will allow you to keep your assessments aligned with the current possibilities introduced by GenAI tools, and include the corresponding adaptations required.
With the constant improvements and frequent updates of GenAI tools, it is important to adopt an iterative approach with your assessments and identify the priority areas where you might make some adjustments. At the beginning of each team, you may want to review the assessment tasks and student performance, to make sure that the tasks are kept meaningful and relevant to student learning.
To help identify the priority areas where you might want to adjust your assessments, we encourage you to refer to the five-stage assessment design process, starting from defining the learning outcomes, planning the course content, developing assessments, delivering the assessments, collecting student feedback, and reviewing the assessments accordingly.
Five-stage assessment design process
The following graphic provides a description for each stage of the assessment design process, and highlights the possibilities of GenAI and adaptations required. You can click on the list below the graphic to learn more about each stage.
Stage 1: Define learning outcomes
Define clear and measurable learning outcomes that describe what your students will be able to know, do, or put into practice by the end of a learning sequence.
Since GenAI tools can now easily perform tasks such as generating summaries, reading notes, and analyzing large datasets, you may need to revise the learning outcomes to emphasize higher-order thinking skills from Bloom’s taxonomy (PDF)—specifically critical analysis, evaluation, and creative synthesis that require uniquely human cognitive abilities.
Stage 2: Plan course content
You could consider designing activities and assessments that train students to use GenAI tools effectively, critically, and ethically throughout the course. This approach can support students to develop both human skills and GenAI-related competencies simultaneously.
Identify and structure your course content, assessments and learning activities, and ensure proper alignment with the intended learning outcomes.
Stage 3: Develop assessments
Create assignments and evaluation rubrics to measure student performance and progress toward achieving learning outcomes.
GenAI tools are transforming assessment in two ways: they can automatically generate quiz or exam questions, feedback, and evaluation rubrics, and they can also become integrated into student work through GenAI-assisted writing, simulations, or reflective exercises. Because of these developments, it can be useful to consider shifting from evaluating only final outputs to assessing the processes students use to get there. Learn more in the “Assessments that Involve GenAI” section of this resource.
Stage 4: Assessment delivery
Implement an assessment process (in-person, online, or hybrid) that supports student learning, equity, and academic integrity.
AI may be used in some cases during assessment activities, such as to provide real-time feedback on student work or to act as an interlocutor in a scenario-based assessments. In-class or online synchronous activities could be considered to ensure assessment integrity and capture students’ problem-solving skills (individually or in groups).
Stage 5: Feedback and review
Gather student feedback and regularly review assessment tasks to ensure their continued effectiveness in evaluating achievement of learning goals. You could also consider including students in the design process as co-creators of assessments.