Program level assessment design and the two-lane approach
Source: Teaching@Sydney
Authors: Adam Bridgeman, Danny Liu, Ruth Weeks
Original source: https://educational-innovation.sydney.edu.au/teaching@sydney/program-level-assessment-two-lane/
Published: 2024-09-12
Source type: institutional guidance
Private backup: the full article text is archived in the private repository at archives/articles/educational-innovation-sydney-edu-au-program-level-assessment-two-lane.source.md. It is not published on the public Quartz site.
Summary
Adam Bridgeman, Danny Liu, and Ruth Weeks argue that generative AI makes program-level assessment design more urgent. Their two-lane approach separates secured assessment of learning from open assessment for and as learning. Lane 1 assessments validate whether students have met course or program learning outcomes at selected progression points. Lane 2 assessments support learning, feedback, and responsible AI use in ordinary coursework. The article’s key move is to make the program, course, major, or specialization the design unit for trustworthy assessment rather than asking every individual assignment to carry the full burden of academic integrity.
Pull quotes
Place secured assessment in the program
“Given the difficulty and undesirability of securing all assessments, it is better to identify where in the program to place these lane 1 assessments.”
— Adam Bridgeman, Danny Liu, and Ruth Weeks
Scaffold AI use in lane 2
“Lane 2, which focusses on assessment for learning, is likely to be where we scaffold and teach students how to productively and responsibly engage with generative AI.”
— Adam Bridgeman, Danny Liu, and Ruth Weeks
Program-level design supports assurance
“However, without program-level design guiding where these secure assessments are deployed, it is very difficult to demonstrate the program-level assurance of learning that is legislatively required to award degrees.”
— Adam Bridgeman, Danny Liu, and Ruth Weeks
Big ideas
Claims
- AI-era assessment needs program-level validation
- In an AI world, assessment should focus on watching students think
- AI-assisted homework requires redesign, not just policing
Key evidence and examples
- The University of Sydney’s two-lane approach distinguishes secured assessment of learning from open assessment for and as learning.
- Lane 1 assessments are mainly program-level, secured, and used to validate attainment of learning outcomes.
- Lane 2 assessments are mainly unit-level, open or unsecured, and used to support learning, feedback, and responsible AI practice.
- The article argues that ubiquitous generative AI makes many traditional assessment products easier to mimic and therefore weaker as sole evidence of learning.
- The authors argue that trying to secure all assessment is undesirable, intrusive, and impractical.
- Program-level design can reduce assessment congestion, avoid repeated unit-level tasks, and make feedback more coherent across a student’s course.
- The article distinguishes this approach from full “programmatic assessment,” which is more common in competency-based fields such as medicine, dentistry, and veterinary science.
Education relevance
High relevance for AI-era assessment design, higher-education policy, school and district assessment planning, academic integrity, and Clay’s assessment work. The source is especially useful because it adds an institutional architecture layer to the wiki’s existing focus on visible thinking, process evidence, and separating AI literacy from assessment integrity.
Durability note
The University of Sydney policy references may age, but the two-lane distinction is durable: schools and universities will still need to separate AI-supported learning experiences from selected validation moments where student understanding must be trusted.