Feedback Quality Analyser & Rewriter

What This Skill Does

Takes a piece of teacher or peer feedback, analyses it against Hattie & Timperley's (2007) four-level feedback model, and rewrites it to improve its level, specificity, and actionability. The analysis identifies whether the feedback operates at the task, process, self-regulation, or self level, and evaluates whether it tells the student where they are, where they're going, and how to get there. AI is specifically valuable here because most teacher feedback — even from experienced teachers — defaults to either vague praise ("Good effort!"), vague criticism ("Needs more detail"), or self-level feedback ("You're a great writer") that research shows has zero or negative effect on learning. Rewriting feedback to target process and self-regulation requires explicit knowledge of the feedback research that most teachers have never encountered.

Evidence Foundation

Hattie & Timperley (2007) conducted a comprehensive meta-analysis finding that feedback has an average effect size of 0.73 — one of the highest in education research — but with enormous variability. Effective feedback answers three questions: Where am I going? (feed up), How am I going? (feed back), Where to next? (feed forward). They identified four feedback levels: task (correctness), process (strategies used), self-regulation (student's monitoring and control), and self (personal praise or criticism). Task-level feedback improves immediate performance; process-level feedback improves strategy use; self-regulation feedback builds independence. Self-level feedback ("You're so clever" / "Disappointing work") has no positive effect and can undermine learning by directing attention to ego rather than task. Kluger & DeNisi (1996) found that one-third of feedback interventions actually decreased performance — typically when feedback threatened self-esteem or directed attention away from the task. Wisniewski et al. (2020) updated the meta-analysis and confirmed that feedback containing information (specific, task-referenced) is significantly more effective than feedback containing only judgments (grades, praise, criticism). Shute (2008) identified that effective formative feedback is specific, timely, non-threatening, and focused on the gap between current and desired performance.

Input Schema

The teacher must provide:

• Feedback text: The existing feedback to analyse. e.g. "Good work! You clearly tried hard on this. Next time, add more detail to your analysis." / "7/10. Some good points but your conclusion is weak."
• Task context: What the student was doing and what the learning objective was. e.g. "Year 10 History essay on causes of WWI. Learning objective: construct a causal argument using evidence."

Optional (injected by context engine if available):

• Student work summary: Brief description of the actual student work
• Student level: Year group and ability profile