Last week, former teacher and Senior Professional Development Manager at The Assessment Network, James Beadle, shared the first part of his two-part series of blogs on the key questions teachers should consider when thinking about AI and E-assessment in the classroom.
In part two, James will continue to combine his assessment expertise and teaching experience to explore some of the other essential questions teachers should be asking in the context of AI and assessment in the classroom.
Will there be any unintended side effects (washback)?
Teachers often feel that they need to adapt their teaching to meet the demands of high-stakes assessments. If an English language qualification assesses writing, listening and reading skills, but does not assess speaking skills, then to what extent should speaking skills be taught within the classroom?
Whilst they may be desirable, if they are not assessed, then they may receive a lower priority, particularly if teachers already feel they lack sufficient time to deliver the curriculum.
These unintended impacts of an assessment on teaching and learning are often referred to as washback.
Whilst washback is often viewed within the context of high-stakes assessments, it is important to recognise that the vast majority of assessments experienced by learners likely take place within a classroom context, and that significant changes to the manner in which they are carried out may have unintended consequences. Many of the assessment activities you currently carry out might involve a significant social element: ‘Think, Pair, Share’ would be a good example of this. In a move to digital, and AI supported assessments, what would the social interactions in your classroom look like? Do students still have opportunities for peer-to-peer assessment, and helping each other learn?
Another unintended impact might be that we end up worsening the inequalities already present within our classrooms. In England, research has increasingly highlighted that there is a ‘digital divide’ between students, with those already disadvantaged most affected.
Consider, will all students be able to access any platform you choose to the same extent? What does it look like on a mobile phone? Are students using a touch screen going to be at a disadvantage compared to those with a keyboard? If we are making significant changes to how we carry out assessment within and outside our classrooms, it is crucial that they benefit the learners who need it most.
There is also a risk that in the desire to ‘personalise’ learning, AI and e-assessment platforms direct learners into different pathways, with each pathway containing varying material at varying levels of difficulty. As Tim Oates discussed in more detail, there is a real risk such a system leads to students downgrading their own personal aspirations, creating a cycle of low attainment. This can then make it very challenging for teachers to address misconceptions and deliver direct instruction across the class. If each student has seen a different question, it becomes far more complex to model a correct solution.
If different students are studying different materials, far more work is required to then bring the class back together to work on a common area. As the next question discusses, it is key that any systems we use, act to reduce, rather than increase, teacher workload.
Will it reduce overall workload?
Aristotle is attributed with the saying ‘Nature abhors a vacuum’: if there is empty space, then something will seek to fulfil it.
The same can often feel true about teacher workload: if something is removed, something else quickly replaces it. In a survey carried out by Teacher Tapp in October 2023, marking students' work was identified as a workload issue by 32% of primary teachers and 24% of secondary teachers.
This is something that could potentially be reduced by the adoption of AI and E-assessment. However, is there then a risk that the time gained is then spent moving assessment data from an external platform to an internal data management system, or producing summaries of this data for middle and senior leaders?
How long will it take you as a teacher to understand the data generated and incorporate it into your teaching?
In the same survey, carried out by Teacher Tapp, recording and analysing attainment information was identified as a workload concern by 29% of both primary and secondary teachers, and it is important that any platforms adopted do not further add to this.
Final thoughts
The questions raised here may seem overly critical, but the intention of them is not to necessarily dissuade teachers from adapting AI and E-assessment platforms. E-assessment has been present in many classrooms for decades now, and AI certainly has a role to play in both improving the quality of assessment and reducing the associated workload for teachers.
At Cambridge, we’ve released guidance on how we are putting people at the heart of the generative AI revolution in an educational context.
Ultimately, we question our students in order to help further their own understanding. It is in this spirit we should also seek to question any AI & assessment platform we seek to use: only by understanding what these platforms can, and can’t do, and how valid the judgements and feedback they give are, that we can ensure that the ones we put in place deliver the excellent outcomes all our learners deserve.
This is part two of a two-part series on AI and E-Assessment Within the Classroom. You can learn more about our latest course opportunities to enhance your assessment practice.
The Assessment Network is part of Cambridge University Press & Assessment. We provide professional development for impactful assessment.