Will AI Marking and Feedback Uplift or Undermine Teaching?

Introduction

As educators strive to provide the best learning experience for their students, the use of AI in marking and feedback has emerged as a topic of great interest. This blog post delves into the potential risks and benefits of integrating AI into the assessment process, examining both the clear advantages and drawbacks, as well as the more nuanced potential risks and benefits that may arise from its implementation. By carefully considering these factors, we aim to provide a balanced perspective on the role of AI in education and its impact on the future of teaching and learning.

Clear Risks

1. Ethical Concerns Regarding Data Privacy and Security

The use of AI raises important ethical questions about data privacy and security, as student data is collected and analysed by third-party systems, potentially exposing sensitive information to unauthorised access or misuse. Ensuring the protection of student data is paramount when implementing AI in educational settings.

Clear Benefits

1. Save Teachers Time

By using AI to streamline the assessment process, teachers will have more time to focus on the myriad other tasks on their plate, such as lesson planning, one-on-one student support, and professional development.

2. More Timely Feedback

AI systems can provide quicker feedback to students, allowing them to receive timely guidance and support. This enables students to act on feedback while the learning is still fresh in their minds, potentially leading to faster progress and better outcomes.

Potential Risks and Potential Benefits

1. Quality of Feedback to Students

Marking and feedback is a powerful tool that enables students to identify their strengths and areas for improvement, promotes self-reflection and metacognitive skills, provides guidance and direction for their learning journey, and offers motivation and support as they progress towards their goals.

The utilisation of AI in this process, if not implemented properly, could compromise these vital benefits. Students may even respond negatively to AI-generated feedback, disregarding it due to a lack of trust or a lack of accountability.

On the other hand, the implementation of AI could also improve the quality of feedback. Teachers are often overwhelmed and struggle to spend hours giving feedback to each student and also keep up with the ever-growing body of literature on best feedback practices. A  AI model could effectively incorporate research-based feedback methods and spend the equivalent of hours providing feedback.

2. Quality of Feedback to Teachers

Grading assignments and providing feedback to students is not only advantageous for the learners but also for the educators. It allows teachers to gain a deeper understanding of each individual student's progress, guiding them in tailoring their support to meet the unique needs of every learner. Moreover, it provides valuable insights into the overall class performance, enabling educators to effectively support the entire group and refine their underlying teaching methodologies and resources.

By removing teachers from the grading and feedback process, these essential benefits may be lost, hindering the growth and development of both students and the teacher.

However, if AI is used as a tool rather than a replacement, it could provide similar services to data analytics tools, enabling teachers to gain even deeper insights.

3. Erosion of Teacher-Student Relationship

When a teacher provides feedback to a student, it facilitates dialogue and helps create a supportive and collaborative learning environment, fostering a stronger relationship between the student and the teacher.

Replacing a teacher with AI could negate this benefit and degrade the teacher-student relationship.

However, AI could also enable teachers to spend more time discussing feedback face-to-face with students, rather than writing it, potentially enhancing the teacher-student relationship. By automating certain aspects of the feedback process, AI can free up teachers' time for more meaningful interactions with students.

4. Increase in Bias and Inaccuracies

AI systems may introduce bias and inaccuracies based on the data they are trained on, perpetuating or even amplifying existing inequalities and leading to unfair or incorrect assessments of student work.

However, if designed and implemented properly, AI can help improve the consistency and objectivity of grading by applying standardised criteria and reducing the influence of human subjectivity and bias. By using diverse and representative training data, AI systems can minimise bias and provide more equitable assessments of student performance.

‍

Finding the Balance: Getting the Benefits Without the Risks

Implementing AI for marking and providing feedback offers significant benefits but also poses risks. Ignoring the technology could mean missing out on its transformative potential, while using it irresponsibly could lead to poor outcomes. Integrating AI is similar to introducing personal devices like iPads or laptops; it is the teacher's responsibility to use the technology effectively to best serve their students.

Determining the ideal integration requires further research. However, currently, an AI-guided approach rather than full automation is likely the best option. This approach may not save as much time as complete automation - perhaps 50-90% instead of 100% - but it balances efficiency and quality. It allows teachers to benefit from AI's insights and speed while maintaining control over the process.

‍