Do this only if the format matches
Confirm the round with recruiting first. This is an original practice exercise, not a claim that a specific company uses this exact prompt.
Protocol
- Run one control and record the exact observation without interpretation.
- Write three competing hypotheses that predict different results.
- Change one factor at a time before testing interactions.
- Use controls or repeats where variance could produce a false pattern.
- Find one boundary where the leading claim stops holding.
- Present the claim, evidence, uncertainty, and next discriminating experiment in five minutes.
Deliverables
- A probe table with predictions and observations.
- One interaction effect and one negative result.
- A bounded behavioral claim, not an architecture guess.
- A five-minute research readout.
Graduation gates
Every claim has a probe that could have falsified it.
Observation and interpretation remain separate.
The readout exposes uncertainty and boundary conditions.
The next experiment is chosen by expected information gain.
Starter files
Assignment READMERules and readout contract.Open file →Query clientUse this interface without opening the oracle.Open file →Probe templateA minimal experiment ledger.Open file →