Using quiz data to plan your next Computing lesson
Quizzes tell you far more than a score. Used well, the results from last week's homework or starter can decide what you reteach, what you skip and what goes on the board when students walk in. Here's a simple way to turn quiz data into next lesson's plan in about ten minutes.
Look past the average
A class average of 72% hides almost everything useful. It could mean everyone knows most of the topic, or that half the class knows it well and half has hardly started. Three views are worth more than the average:
- Question by question: which questions did most of the class get wrong?
- Topic by topic: which topics are strong across the class, and which are still developing?
- Student by student: who is behind on several topics, and who hasn't attempted the work at all?
Step 1: find the questions most of the class got wrong
Sort the questions by how many students got them right. The two or three at the bottom are where to start. Then read the wrong answers, not just the right one. If most students chose the same wrong option, that's a shared misconception and worth reteaching to the whole class. If the wrong answers are spread out, the question may just have been hard, or worded badly. See 12 common misconceptions in GCSE Computer Science for the ones that come up most.
Step 2: decide whether to reteach, revisit or move on
| What the data shows | What to do next lesson |
|---|---|
| Most of the class got a question wrong, with the same wrong answer | Reteach it to everyone: a short explanation, a worked example, then a fresh question to check |
| About half the class got it wrong | A quick reteach, then put the question in the Do Now again next week |
| A few students got it wrong | Move on, and support those students individually or with targeted homework |
| Almost everyone got it right | Move on, and bring it back in a few weeks so it isn't forgotten |
Step 3: build the Do Now from the gaps
The quickest way to act on the data is next lesson's starter. Take the questions the class found hardest, add one from an older topic, and put them on the board as students arrive. Go through the answers straight away, focusing on why the common wrong answer is wrong. See How to run a Do Now in Computer Science.
Step 4: check whether it worked
A gap isn't closed just because you've retaught it. Ask the same kind of question again a week later, and again after a few weeks. If the class still gets it wrong, try a different explanation rather than the same one again. This is retrieval practice doing its job: see Retrieval practice in Computing.
Use topic data for longer-term planning
Question-level data shapes tomorrow's lesson; topic-level data shapes the next few weeks. Before a mock or the start of revision, look at which topics are still developing for most of the class, and plan to revisit those first. For individual students, a topic-by-topic view makes it easier to give specific advice than "revise more".
Keep it to ten minutes
- Look at the same few things each week: the hardest questions, any topic most of the class is still developing on, and who hasn't done the work.
- Write down one action per class, such as "reteach the order of volatility" or "Do Now on binary shifts".
- Don't try to fix everything at once. Two or three gaps acted on properly beat ten noted and forgotten.
How RADevo helps
RADevo does the sorting for you. It shows each class's hardest questions, and Knowledge Profiles show every student's progress topic by topic as Strong, Secure or Developing. The Do Now builder starts from the questions your class got wrong most often last week, so the starter is ready in a couple of clicks. On the AI Analysis plan, you can also get a short written summary of a class's strengths, the topics to revisit and suggested next steps. Students see the same picture in their own Topic Map, with a best next step, and every question they get wrong comes back in their Revise today. Try the free revision quizzes to see the questions behind the data.