This tool gives you a short, original investigation with its variables and data. You choose the main weakness, choose the most suitable improvement, and write a sentence or two of reasoning that uses the evidence. Then it shows a model critique and explains why the other improvements do not fix the problem.
All four scenarios are safe classroom-level examples.
How do I use it?
- Choose an investigation: light and plant growth, the mass of a metal block, reaction time, or warming a cup of water.
- Read the description, the variables and the data table. Note what was changed, what was measured and what should have been kept the same.
- Pick the main weakness from the list: uncontrolled variable, systematic error, too small a sample, or resolution too coarse.
- Pick the most suitable improvement from the three options offered.
- Write your reasoning. Quote at least one value or variable from the data. A sentence or two is enough, and it must not be empty.
- Press Get critique. Reset goes back to the starting scenario.
Example walk-through
The tool opens on the mass of a metal block. The same block was weighed three times on one balance and gave 78.0 g each time. The balance shows 2.0 g with nothing on it, and a known 50.0 g mass reads 52.0 g.
Choose “systematic error (calibration)” as the weakness. For the improvement, try “repeat the reading ten more times and take the average”. Write: “All three readings are 78.0 g, but the empty balance shows 2.0 g.”
The tool confirms the weakness matches the model critique, then says the improvement does not fix it: repeating reduces random scatter only, and a biased balance gives the same 2.0 g error every time. The option that does fix it is to zero or calibrate the balance with the known mass, or subtract the known offset. The true mass of the block is therefore 76.0 g.
Try the other scenarios too. The plant growth one rewards you for spotting unequal watering, and the warming-cup one for noticing that a 1 °C difference equals the smallest mark on the thermometer.
How do I read the result?
- Weakness line: whether your choice matches the model critique, and what the model classes it as if not.
- Your improvement: whether it addresses the weakness, with a reason.
- Model critique: the authored explanation linked to the numbers in the table.
- Other alternatives: all three options, each with why it does or does not help.
- Tip line: appears if your reasoning had no digit. Add a value and try again.
What are the limits?
The critique is authored for these four scenarios only, and the tool does not judge your own wording. It is useful as a prompt to compare your reasoning with a model answer.
It gives no instructions for hazardous experiments. Never attempt any investigation without proper supervision and a risk assessment from your school. Check your syllabus on the Cambridge subject page for what your course expects in practical-style questions.
Which lessons explain the output?
Read identifying a controlled condition across linked observations, describing a safe conceptual measurement improvement and why repeated readings do not fix a biased method.
Then see comparing an experimental conclusion with an alternative. The topic page is investigations and evidence, with a mixed practice set.
For the variables side, try identifying confounding variables in a mixed scenario. The tools page lists the rest.
If you understand the ideas but evaluation answers still feel vague, our team can practise them with you. See online one-to-one Combined Science tuition.