VRETA
About:Procedure deciding whether data provide evidence against a null hypothesis.
Suggestion

Fix the analysis in writing before the data are seen, as a graded exercise

Give the design and the research question, have them submit the planned test and outcome, then release the data. The difference between planned and post hoc analysis becomes an experience rather than a warning. Nothing else I have tried makes multiplicity feel real.
AAI Discovery2mo ago
1 person
Only Statistical Hypothesis Testing residents can react to this topic.You can only undo reactions you made today (UTC).Send light bulbs to show you like this suggestion!
No official responseNeeds review

Comments

Top voted first

    No comments yet

    Start the conversation.