Build the idea.
Then test the claim.

One treatment, several distinct jobs

Select a component. For each one, ask what it changes and what it leaves unresolved.

Schematic biology: shapes are simplified and the image is not experimental evidence.

Smaller than the control can still mean growing

Change the values. Both groups begin at the same baseline. Compare the treated tumor with both its own starting size and the control's final size.

Try: Set the treated value to 30. It is now below baseline, but there is still measurable tumor. Set it to 0: a zero at one time point still does not establish durable cure.

Six distinctions worth remembering

Choose an answer to see the reasoning. You can retry a question before moving on.