A replication of Oakes (1986) on misinterpretations of p values and et al (2014) regarding misinterpretation of CIs. (open access: cambridge.org/core/journals/) It's always good to be replicated, but I'd like to put it in a context. 1/
Conversation
When we did the 2014 modification of Oakes to ask about CIs, one of our intents was to see whether the argument CI advocates were using about p values also applied to CIs. It turns out it does (and this paper reinforces that). What does this mean? Could mean one of sev. things 2/
2
1
4
You can't conclude that because your study asks different questions for p-values and CIs.