Papers
Reading a paper is not a front-to-back exercise.
Four blocks first
Abstract for the claim, methods for sample and equipment, results for whether the numbers support it, limitations for what the authors admit.
- Check sample size and scanner model against your setting
- Look for a control group; without one, discount the claim
- The limitations section often matters more than the conclusion
How to record
Three lines per paper: what was done, what was found, what it changes for you. If you cannot write them, you have not finished reading.
Findings depend on study design; do not transplant them into practice directly.