Results and feedback
Read exam results without jumping to conclusions
Exam analytics are a prompt for investigation. A low score can reflect a learning gap, an unclear question, a time problem or an access issue. Read the group pattern and the question details together before making a decision.
Begin with who actually took the exam
Check the number of started and submitted attempts and whether repeated attempts are included. A group average based on a few early submissions is not yet a class conclusion. Compare like with like: the same exam version, scoring rules, attempt policy and availability window. If any of these changed, make that limitation explicit when sharing results with colleagues or participants.
Look beyond a single average
The average summarises the group but can hide very different experiences. Inspect the median and distribution where available. Two groups can share an average while one is clustered around it and the other contains two separate bands. Check whether the passing threshold was chosen for a clear purpose. A pass rate is a count against that threshold, not an independent measure of question quality.
Use question patterns to form hypotheses
A low correct rate on one item is a reason to examine that item and its learning goal. Read the wording and verify the key, then inspect common wrong answers. Several questions on the same skill can support a stronger interpretation than one isolated result. Treat short timings and tab-switch counts cautiously; neither proves guessing or misconduct by itself.
Close the feedback loop
Choose a small intervention: explain a misconception, model a worked example or provide targeted practice. Run a comparable follow-up check and look for a change in the relevant skill, rather than only a higher overall score. When exporting results, share only what the recipient needs and define each column. Keep identifiable participant data out of public presentations and shared examples.
Before you publish
- Attempt counts and repeated attempts are understood.
- Unusual questions have been checked against their answer keys.
- One specific teaching action follows from the analysis.
Questions, answered
- Does a low average mean the exam was too hard?
- It is one possible explanation. Check topic coverage, instruction, timing, access and question quality before changing the standard.
- Can two exam averages be compared?
- Only with context. Different questions, groups, attempts and scoring rules can change the average even when understanding has not changed.
