It Takes a Village... Program-Wide Approaches to Redesigning Assessment in a Time of Generative Artificial Intelligence (GenAI)

Synthesis: Nicola-Richmond et al. (2025) present a qualitative interpretivist study — semi-structured interviews with academics (n = 12) and students (n = 17) at a large Australian university — exploring lived experiences and views of Generative AI in relation to Assessment. Thematic analysis surfaced three themes: (1) emotions, academic role and identity; (2) assessment must change; (3) the work and expertise required to change assessment. Academics and students reported both positive and negative emotions toward GenAI, and pervasive concerns about the complexity and resources needed to change assessment. In response the authors propose that redesigning assessment in the GenAI era takes a village — collaborative, program-wide approaches that bring together diverse expertise (assessment design, GenAI applications, subject matter, program knowledge, industry, and evidence base) to redevelop assessment collaboratively.

Key Findings

• Theme 1 — Emotions, academic role and identity: GenAI evoked a range of positive and negative emotions in both groups. Some academics found GenAI's impact "hurtful" and feared the "diminishment of critical thinking skills" and "the future of my discipline," while others were excited and moved from seeing GenAI as "a threat" to a tool students must learn to use. Academics felt "behind the eight-ball," perceived their role would change (from transmitter of knowledge toward guiding the process), yet valued the continued importance of human interaction.
• Theme 2 — Assessment must change: Both groups agreed Assessment must change to (a) assure student learning (verifying graduates achieve promised learning outcomes) and (b) prepare students for a GenAI world of work. Most opposed banning GenAI, and participants emphasized assessing differently — building GenAI use into the curriculum, designing tasks that minimize unhelpful GenAI use, and requiring critical evaluation of AI output rather than passive consumption.
• Theme 3 — The work and expertise required to change assessment: Academics cited inflexible processes and long lead times, the complexity of redesign (especially in large-cohort, accredited courses), significant time and cost burdens, and the need for pragmatism. They strongly expected universities to provide workload capacity, training, exemplars, and tools.
• The "village" proposal: Effective assessment redesign in the GenAI era exceeds any single academic's capability; it requires a collaborative team with expertise in assessment design, GenAI applications, subject matter, program knowledge, industry use, and the evidence base.
• Program-wide approach: The authors advocate a program-wide (rather than individual-task or programmatic) approach to assessment redesign, allowing strategic placement of GenAI-literacy development and assurance-of-learning points across a qualification while managing resource constraints.
• Universities' role: Institutions must make sustained investments — policy change, frameworks, professional development, and tools (portfolio/curriculum-mapping software) — to meaningfully support academics undertaking assessment change.

Study Design & Method

This is an empirical qualitative study using a qualitative interpretivist approach. Data were collected via online semi-structured individual interviews with 17 students and 12 academics across the university's four faculties (Health; Science, Engineering and Built Environment; Business and Law; Arts and Education). Interviews explored participants' understanding and use of GenAI tools, the impact on assessment practices, the role of the university, how (if at all) assessment needed to change, and the future of university assessment. Data were analyzed using Braun and Clarke's six-phase thematic analysis with inductive coding, inter- and intra-rater reliability checks, and refinement through team discussion and consensus. Ethical approval was granted by the Deakin University Human Ethics Committee (HAE-23-057). The paper focuses primarily on academic data, using student data as a supplement. Limitations include a single-institution sample, minimal demographic data (to protect anonymity), and a single point-in-time data collection amid rapid GenAI change.

What this means for practice

• Instructors. Stop debating bans and redesign instead: most of the 12 academics and 17 students interviewed opposed prohibiting GenAI and favored building its use into the curriculum and requiring critical evaluation of AI output.
• Instructors. Assess differently on two fronts at once — assure that graduates achieve the promised learning outcomes, and prepare students for a GenAI world of work — rather than assuming existing tasks still do either.
• Faculty developers. Convene the "village": assessment redesign in the GenAI era exceeds any single academic's capability, so assemble a collaborative team with expertise in assessment design, GenAI applications, subject matter, program knowledge, industry use, and the evidence base.
• Administrators. Supply the workload capacity, training, exemplars, and tools (portfolio and curriculum-mapping software) that academics said they need, and use policy and framework change to remove the inflexible processes and long lead times that block redesign in large-cohort, accredited courses.
• Designers. Plan program-wide GenAI-literacy development and assurance-of-learning points across a qualification rather than task by task, so strategic placement absorbs resource constraints instead of multiplying them.

Limitations

• Data come from one Australian university, and the authors caution that application to other settings must be considered with caution.
• The sample is small (12 academics and 17 students) and recruited through direct email and learning-management-system invitations from four faculties of one university, with minimal demographic data collected to protect anonymity — which the authors note limits attribution of findings to particular disciplines or areas of practice.
• Data were collected at a single point in time during rapid GenAI change, so participants' use may already have changed significantly since; the authors call for longitudinal, multi-institution work.
• The paper is built on academic data and uses student data only as a supplement, so the student perspective carries less evidential weight than the framing suggests.

Citation

Nicola-Richmond, K., Dawson, P., Partridge, H., & Macfarlane, S. (2025). It takes a village. Program-wide approaches to redesigning assessment in a time of generative artificial intelligence (GenAI). Journal of University Teaching and Learning Practice, 22(7). (CC BY-ND 4.0.)