1. Connection to realityA claim should concern observable objects, processes or testable consequences.
2. TestabilityIt should be clear what supports the conclusion and what could show it to be wrong.
3. Reproducibility of the argumentAnother competent researcher should be able to reconstruct the assumptions, reasoning and result.
4. Independence of scientific judgmentWhen a result conflicts with external interests, what actually receives priority must be distinguished.
5. Evidence above authority and consensusPosition, prizes, journal prestige, school and number of supporters may be social indicators, but they are not proofs.
6. Work with errorsScientific practice searches for and corrects errors; systematic refusal to correct a testable error moves activity toward a pseudoscientific function.