The method
- Find the relevant observation.
- Ask what follows with the fewest added assumptions.
- Eliminate answers that introduce an unmentioned cause or overstate certainty.
Worked example
A library counted more returns after installing an outdoor slot. It did not record who used it. Can we conclude new borrowers joined?
- The measured outcome is returns.
- Existing borrowers could explain the increase.
- The evidence supports increased returns. It cannot identify new borrowers.
Work through a second example
“When the side gate closed for a week, foot traffic at the main gate rose. The counter covered only the main gate.” What can the record tell us about visitors?
- The counter measured entries at one gate. Its higher reading supports an increase there.
- Visitors who normally used the side gate could have switched entrances. The total number entering the grounds could even have stayed the same.
- Choose the narrow conclusion: more entries were recorded at the main gate. A claim about every entrance combined requires additional evidence.