Bayesian Way | NEJM Evidence

[@michaelg4919]

Amazing! Thanks for this short video. I am a big fan of your approach to summarizing science (like abstracts, statistics etc.) in an understandable yet concise way. Hopefully, one day, I'll be able to publish a paper in your journal :)

[@Nephtales]

The best explanation ever

[@julyisinthesummerhere]

this script is amazingly written! great job!

[@NitaJain]

What if both the true difference and the data are variable? What approach is best in that case?

[@haonansun3355]

Well done!!! I enjoyed this video a lot!

[@GardenGuy1942]

I didn’t. It makes me rather upset. They need to TALK LOUDER.

[@anthonyrimicci7858]

Wow! That was great.

[@Godeau03]

Playing the word game on the concept of 'true difference'. In freq approach, the true difference is defined as the mean of the distribution of the difference, so it exists. For example, the height difference between male and female in Duke university, that's not random. One can sample 100 male and female randomly and estimate such true average difference. It's true but more importantly it exists and does not change. So when you ask what if there is no single true difference, it does not stir things up. One can only talk about a concept when it is well established/ defined. The Bayesian prior is also defined on the mean of the distribution of the difference, not on any difference as the video vaguely implied. The criticism of applying Bayesian approach in estimation (such as RCT) is also based upon this argument: how can you define a probability distribution when your parameter of interest is defined as a constant.

[@joaopedrorocha5693]

That's amazing!

[@TiagoPereira-hm1nq]

Bravo!!!👏👏👏👏

[@Polodus]

made my day!

[@delhitamil]

Enjoying Video

[@jackwalsh1468]

Hydrogen to fuel internal combustion engines. Zero Carbon Dioxide and only water vapor. This hits all the environmental check boxes and Net-Zero Carbon.. Even, Greta can't argue against Hydrogen power, logically