Worksheets and research resources · Page 34

RoB 2 trial risk-of-bias guidance

Find the official risk-of-bias tool and guidance for randomized-trial results, with versions for parallel-group, cluster, and crossover designs.

Download free appraisal checklists for several study types and use the accompanying guidance to structure an evidence discussion.

Read openly available lessons in the Unix shell, version control, and research programming to improve reproducible analysis workflows.

Compare providers by relevant experience, written scope, reproducible deliverables, and revision terms.

Record evidence and unresolved questions when comparing statistical support. No invented scores or provider rankings.

Document pairing, correlation assumptions, and sensitivity checks before reconstructing a change-score standard deviation.

Explore cluster size and intracluster correlation without mistaking an effective sample size for a power calculation.

Plan paired differences, limits of agreement, and acceptable error before comparing two measurement methods.

Check the pretest estimate and test context before converting diagnostic likelihood ratios into probabilities.

Connect screening, allocation, treatment receipt, and analysis counts without mixing overlapping reasons.

Detect duplicate keys and unintended row multiplication before joining participant, visit, and laboratory data.

Resolve ambiguous formats, time zones, partial dates, and interval conventions before deriving durations.