The AI-Resistant Shift: More Than Just Outsmarting the Bots
For African higher education institutions, the GenAI challenge is uniquely complex. Educators must balance academic integrity against a backdrop of constrained digital infrastructure, large student cohorts, and distance-learning models.
Rather than relying on unreliable AI-detection software, forward-thinking institutions are redesigning assessments to be inherently human-centric. These formats include:
- Project-based tasks and case studies that require students to apply theory to real-world scenarios.
- Reflective journals and e-portfolios that document a student’s personal, iterative learning journey over time.
- Collaborative tasks and oral presentations that put interpersonal reasoning and live dialogue at the centre of the grading process.
While these formats successfully protect academic integrity, they also demand a completely different style of preparation from students. To understand how students navigate this shift, Prins and Bell (2026) surveyed undergraduate distance learners to map out exactly where they turn for help – and what actually helps them succeed.
The Paradox of Convenience: What Students Use Versus What Works
The study revealed a clear misalignment between the frequency of resource use and the actual educational value those resources provide.
The Resource Usage Hierarchy
When preparing for assessments, students naturally default to convenience. The study found that foundational, easily accessible resources were used most frequently, while active and social resources were largely neglected:
- Module Materials and General Online Content: The absolute go-to resources for the vast majority of students.
- Online Applications and eLibrary Databases: Used far less frequently, often due to perceived complexity or a lack of search confidence.
- Lecturer Consultations and Peer Discussions: Ranked as the least-utilised resources – a common challenge in distance-learning environments where social interaction is not built into the daily routine.
The Real Drivers of Learning Impact
Using Kruskal-Wallis H tests, the researchers analysed whether frequent resource use actually translated into self-reported learning benefits. Surprisingly, the most popular resources – module materials and general internet searches – had no statistically significant relationship with any major learning impacts.
Instead, the most powerful learning benefits came from the least-used, more active resources.
| Learning Resource | Usage Frequency | Statistically Significant Learning Impact | Statistical Significance (Kruskal-Wallis) |
| eLibrary Databases | Low | Improved research skills | |
| Online Apps/ Websites | Medium-Low | Improved research skills | |
| Peer Discussions | Low | More meaningful peer discussions | |
| Lecturer Consultations | Low | More meaningful peer discussions | |
| Module Materials | High | None detected | Non-significant |
This table highlights a critical lesson for educators: simply making resources available is not enough. The strategic value of a resource lies in its capacity to trigger active, high-value learning experiences, which often require more effort from the student.
The True Driver of Satisfaction: The Power of Peer Connection
Perhaps the most striking finding of the study is that no single resource directly influenced student satisfaction. Whether a student spent hours reading module guides or searching the web had zero direct impact on how satisfied they felt with their assessment experience.
Instead, the researchers discovered a mediated pathway to satisfaction. Out of all the learning impacts examined – including applying theory to practice, connecting concepts and managing time – only one was statistically linked to overall satisfaction: engaging in more meaningful discussions with peers ().
In practice, this means that when students use social resources (like peer discussions and lecturer consultations), it enhances the quality of their peer interactions. It is this qualitative shift – feeling connected and intellectually engaged with their peers – that ultimately drives their satisfaction with the assessment. For distance learners who often study in isolation, this sense of connection is incredibly powerful.
Strategic Takeaways for Higher Education
For lecturers at tertiary institutions, instructional designers and academic leaders, this research provides a clear roadmap for supporting students through the transition to AI-resistant assessments.
- Design for Social Connection: Since peer interaction is the ultimate driver of satisfaction, social elements must be intentionally designed into distance learning. Educators should build structured peer-review stages, collaborative online forums and virtual study groups directly into the assessment workflow.
- Scaffold High-Impact Resources: Because students default to basic module guides, institutions must actively guide them toward more complex tools. This means integrating eLibrary search training and specialised digital applications directly into assignment briefs.
- Move Beyond Content Delivery: Providing a library of PDFs is no longer enough. The role of the modern university in the age of GenAI is to facilitate active, collaborative and human-centred learning experiences that technology cannot replicate.
Ultimately, humanising assessment in the era of AI is not just about preventing cheating. It is an opportunity to redesign higher education around what truly matters: critical thinking, scholarly research and meaningful human connection.
Reference List
Prins, R. and Bell, B. (2026) Reimagining Assessment in African Higher Education: How AI-Resistant Formats Shape Student Resource Use, Learning Impacts, and Satisfaction. Unpublished paper presented at the 19th International eLearning Africa Conference, Accra, Ghana, 3-5 June 2026.
Full references are available in the original manuscript.
Frequently Asked Questions: AI-Resistant Assessments & Higher Education
1. What are AI-resistant assessments in higher education?
AI-resistant assessments are human-centric evaluation formats designed to measure higher-order critical thinking and authentic learning rather than mere information recall, making them difficult for Generative AI (GenAI) tools to replicate. Examples include project-based tasks, reflective journals, e-portfolios, and collaborative oral presentations that focus directly on the learning process and live interpersonal dialogue.
2. What is the “Paradox of Convenience” in distance learning resource usage?
The Paradox of Convenience describes a distinct misalignment where distance-learning students frequently utilise easily accessible, foundational resources (such as module materials and general web content) that have no statistically significant impact on actual learning outcomes. Conversely, high-impact resources that require more effort and active engagement (such as eLibrary databases, peer discussions, and lecturer consultations) are used far less frequently despite driving significant academic benefits.
3. What is the primary driver of student satisfaction in online and distance education?
Empirical research indicates that static content delivery and individual study materials do not directly influence student satisfaction. Instead, the primary driver of overall satisfaction is meaningful peer connection and intellectual engagement. Engaging in collaborative tasks and peer discussions helps eliminate the isolation inherent in distance learning, significantly boosting the student experience.
4. What research did the IMM Graduate School present regarding AI-resistant learning?
Researchers Riana Prins and Brad Bell from the IMM Graduate School presented a pioneering study at the 19th International eLearning Africa Conference in 2026. Their work analysed how undergraduate distance learners engage with AI-resistant assessment formats, mapping out the clear discrepancy between the convenience of resource use and actual educational value, while providing a framework for modern curriculum design.
5. How does the IMM Graduate School design its distance-learning assessments to counteract GenAI cheating?
Rather than relying strictly on unreliable AI-detection software, the IMM Graduate School focuses on humanising its evaluation models. The institution leverages a strategic framework that incorporates process-orientated e-portfolios, collaborative oral tasks, and structured peer-review cycles directly into assignment rubrics to ensure academic integrity in the age of artificial intelligence.
6. How does the IMM Graduate School support student connection and reduce isolation?
Grounded in data showing that peer interaction drives student success, the IMM Graduate School intentionally “engineers social contact” within its academic programmes. This is achieved by embedding facilitated online forums, virtual study groups, and interactive assessment workflows to ensure distance learners remain highly connected, collaborative, and intellectually engaged throughout their studies.