Adjusting self-assessment with AI feedback: The core purpose of self-assessment
Self-assessment and reflective practice stand as fundamental pillars of effective learning, serving as a necessary bridge between students’ experiences and the meaningful knowledge they derive from them. The education literature consistently emphasizes that reflection is not just an auxiliary task, but a central feature of experiential education aimed at solidifying the connections between what students have done and the learning they have gathered. Unless students reflect deeply on their experiences, it is impossible to be certain of the learning outcomes and meanings they ascribe to the process.
Involving students in self-assessment activities is of great educational importance as it actively promotes the skills necessary for continued growth and academic maturation. [1]. These activities promote reflective practice and self-monitoring abilities, which are prerequisites for developing independent learning. When students are able to accurately judge their own performance, they take ownership of their own learning process, are more motivated, and develop an important range of personal and transferable skills.
Post-experience reflection asks students to reconsider their perceptions, assumptions, and understandings in light of what actually happened. This means challenging simplistic conclusions, considering alternative perspectives, and comparing current insights with previously formed insights. True critical reflection requires reviewing existing knowledge, questioning assumptions, and forming new perspectives. When students connect these insights to broader academic ideas, they engage in the kind of analytical rigor that defines deep learning.
In this article…
The fundamental issue: Why do students struggle to reflect accurately?
Despite the value of self-assessment, the process is often undermined by psychological and cognitive barriers that make subjective reflections unreliable. Students have a natural tendency toward self-improvement, which creates a positive bias in their self-evaluation. They focus on what went well and often overestimate their performance to avoid negative feedback.
Many students believe that they view their work objectively, so there is little reason to adjust their judgment. This false sense of accuracy becomes a self-reinforcing cycle. Research shows that the more confident students are in their performance, the less they will accept external correction. As self-esteem increases, sensitivity to feedback decreases.
Traditional feedback typically arrives after a task is completed, so it’s too late to challenge exaggerated perceptions that have already been formed. External factors unrelated to performance can also distort self-evaluation, such as cultural norms about modesty or fear of embarrassing yourself in front of colleagues. [2]. These influences can cause students to undervalue their work or exaggerate it to protect their egos.
The purpose of reflection is to enable students to realistically judge their own performance. This is where technology becomes essential. AI helps adjust students’ internal perceptions using objective external performance data by providing depersonalized, data-driven feedback at the moment reflection occurs.
Figure 1: Overcoming subjectivity in self-evaluation
The role of AI: Introducing objective external feedback
artificial intelligence [AI] supports more accurate self-assessment by providing students with immediate, specific feedback without personal judgment. When used responsibly, AI can act as a sounding board that enhances the work of both students and educators, rather than replacing human judgment. AI-powered feedback also captures details that are often missed by traditional methods. Measure speech patterns, coherence, structure, and use of filler words. Analysis tasks can show how students’ assumptions and biases shaped their conclusions. This timely data helps prevent subjective impressions from becoming fixed.
A strong teaching strategy involves shifting critique towards the output of the AI. For example, instructors can ask students to generate an AI analysis of a passage of literature or a market trend. Students then write their own analysis, critiquing the AI version and pointing out where nuance is missing, bias is introduced, or historical or cultural context is ignored. This takes the pressure off of protecting your job and allows you to make important decisions in a low-risk environment.
To support academic integrity, educators must clearly articulate which AI tools are allowed and how they should be used. Students can then submit a short testimonial explaining how the AI shaped their thinking and how far their reasoning has progressed. This turns everyday practices into meaningful evidence of learning and discernment.
Structured reflection: A framework for turning data into insights
Objective AI data is most useful when it fits into a structured reflective framework that takes students from simple recall to actual analysis and action. Borton’s development framework is a clear way to do this, built around three stages:
what
Instead of memory, students use AI-generated data to describe their experiences. Metrics such as articulation score and filler word count provide an accurate starting point and remove uncertainty about what happened. so what
Students interpret the data. AI reveals the gap between perception and performance, prompting you to consider why an outcome occurred and how your assumptions shaped it. what now
Students put these insights into action. AI highlights specific weaknesses, making it easier to set focused next steps and build a steady cycle of improvement.
Using AI within this structure turns self-assessment into a focused loop. Students work on several priorities at a time and make steady, meaningful progress based on objective data.
Figure 2: Application of the structured reflection framework (Borton adaptation)
Practical application: A guide to reflection in high-stakes scenarios
AI-powered reflection is especially useful in high-stakes situations, such as academic or professional interviews, where students must demonstrate clear reasoning, strong storytelling, and understanding of institutional values. In competitive fields such as medical school admissions, candidates are expected to connect personal experiences to complex medical issues and broader policy ideas. AI tools can help students meet these expectations by aligning their performance to the standards of top programs.
The AI simulator provides a safe space to practice and refine your answers. After each session, the technology analyzes responses across key areas such as structure, clarity, pacing, and use of filler words. As questions become more complex, AI points out gaps in the narrative, such as weak links between personal stories and policy discussions. Mock interviews can also expose students to ethical dilemmas, local policy contexts, and equity-focused scenarios that encourage deeper, more thoughtful reflection.
In fact, students can also compare their grades to expectations outlined in official school guidelines, such as those detailed in this guide for Harvard Medical School. This will help you see whether your answers are aligned with your institution’s values and priorities, and provide specific areas for targeted improvement.
AI can also support insights in areas that require policy understanding. Policy overview mode allows you to quickly summarize key frameworks such as the Massachusetts Model for Cost Containment and the Mental Health ABC Act. Students can use these summaries to practice writing clear, solution-focused responses. By critiquing the AI synopsis and identifying where it oversimplifies or misses local context, we can become more discerning and gain a deeper understanding of the policy landscape that needs to be understood.
Actionable steps for students: Build reflective habits
To harness the objective power of AI and build strong habits of self-reflection, students must adopt a systematic approach that emphasizes low-stakes, frequent engagement, and structured analysis.
Request proofreading now
Start each session by running one scenario through the AI to get an objective baseline. Review your data now while the experience is still new. criticize the machine
For complex tasks, ask the AI for example answers and critique them. Please point out where I’m missing nuance or context. This builds critical thinking without feeling pressured to criticize yourself. follow framework
Use clear reflection models. You start with what (the AI data), move on to what (why the score occurred and what assumptions you made), and now end with what (one specific action for the next attempt). Prefer one change
Focus on one AI-flagged improvement at a time. Trying to fix everything at once weakens the feedback loop. Track your progress
Keeps a simple log that compares your self and AI scores. Notice how the data challenged your initial impressions and how your understanding of the task changed over time.
Conclusion: Developing insight and ownership for self-assessment with AI
Using AI for self-assessment allows students to judge their own performance more realistically. Immediate, objective data disrupts normal patterns of self-enhancement and bias, making reflection more accurate and useful. The purpose is not to replace reflection, but to enhance critical thinking. AI guides the process, but students remain responsible for their judgment and ethical choices. As AI feedback becomes routine, students will develop stronger self-monitoring and analytical skills that support both academic growth and long-term professional development.
References:
[1] self-evaluation
[2] Accuracy of students’ self-evaluation: Research directions and points to note
Image credits: Images/tables within the article text were created/provided by the author.
automation
The ultimate no-code web scraper, bot builder, and data extraction tool for efficient and hassle-free tasks.
