Fascinating.
Nicely done video series on experimental design and analysis. The series also covers reporting.
Experimental design: Experimental Design - YouTube
Analysis and reporting: Analysis and Reporting - YouTube
Given the extensive criticism of so many ex fizz papers and the recent discussion of science, I thought this might be of interest.
Introduction at Reproducibility in Research: An Introduction - YouTube. Might even be useful for CME credit through Indiana Univ School of Medicine.
Fascinating.
He gets p-values right!! Hard to find that in this day and age
Analysis and Reporting Module 2b: P Values - YouTube
Agreed.
The next video explains that even with typical p-values and powers, we can have a false discovery rate (seemingly significant results that were not true effects) of 50% or more. This is amazing and frightening. Analysis and Reporting Module 2c: P Value Meaning - YouTube
Obligatory xkcd: xkcd: Significant
You can try it yourself!
What color jelly bean causes (or prevents!) acne?
In most cases there is only one statistical measure that really matters:
P(H>U) >= T ?
The Significance level of P is determined by whether it exceeds T
"What is the probability that your statistics will be reviewed by someone who regards himself as a greater statistical expert than you?"
How much criticism can you Tolerate.
Cynical...I know.
Didn't watch yet, but I prolly will, cuz I'm a nerd. But do they talk about effect sizes?