Oh I apologize. I did not realize you were asking for a critique of PS. I have Pearl's book but haven't read that far yet. I'll check it out.
anything from Paul Rosenbaum for an appropriate take on PS.
I don't know that I dislike it as much as you, but I do appreciate say, books that provide arguments that have been made. For and against. A lot of philosophy and history books do this. You decide which one is more convincing.
I've looked through quite a lot of Bayesian Statistics books. Hard to find one where they don't talk about Frequentist stats and how they find it the lesser option. Ive only found one Frequentist type book where the book(Aris Spanos) felt dogmatic. I kinda liked seeing it >:)
I've tried both my personal and student emails to register and I keep getting an error about my email not being a valid email.
Perfect
Not p-hacking but definitely some sort of cheating verb.
Do you talk about the arguments for NHST?
psycnet.apa.org/record/1997-...
Confidence Intervals are inverted hypothesis tests.... p-values are fine.
Wikipedia is good. But I found two Statistics Encyclopedia's that do what I want from Wiki better at times.
Not a Bayesian but maybe you could also add what your goal is? For example are you interested in applied Stats or Math/theo Stats. Im interested in the latter so I'm going to read this one at some point
www.amazon.com/dp/038796098...
Register here (for FREE): www.causalscience.org
Can't join? Subscribe to get a notification when the recording is out: causalbanditspodcast.com
**Read the report: genderedinnovations.stanford.edu/case-studies...
6/6
#CausalSky
Laughs in C++
The third installment of the “how should we actually construct our causal graphs anyway” series is out now! 👇🏼
Nick & I ask the question: can we just get an LLM to tell us what belongs on the graph?
I have a few of Neyman's papers saved. Reading about his inductive inference is interesting.
That is surprising to my noob stat brain.
you if all you have is college algebra. But I think it's a great work book to get your feat wet! I've read both the book and workbook in it's entirety and I think the authors have written one of the best HS stat books.
There are two books I think you can still use. Statistics by Freedman and Mathematical Methods in Statistics by Freedman is a solid choice if you want to learn about Statistics and some probability. The latter will challenge..
www.amazon.com/Statistics-F...
www.amazon.com/Mathematical...
I think Statistics and Probability courses are worth the effort if it is calculus based probability. So, if you were in HS and had to choose between AP Stats or AP Calculus. AP calculus is worth more of your time. Despite that, if you really don't plan on getting up to Multi variable Calculus...
I can grant that for now. Why do you suppose that is? I would imagine the one with a higher mathematical exposure would have the better understanding. Or from a Philosophical one.
Would you say the same thing if they went from Bayesian to Frequentist? If yes, then it seems like an advantage independent of what the switch is to. I'm a BS in Statistics, and I know that interpretation is wrong and why it is wrong. This is a critique of the education system.
That's frustrating but I see your point. That individual is equivocating on the word probability 🥲
They preface it with a promise of "this is what you want to know anyway". But imagine if we did that with say, calculus. A student doesn't quite understand integrals. Would we suggest them just do something else? As a stats/math tutor, I find this a bad solution.
"I see too many people casually using Bayes as just another stat method"
I'm a student, so I don't really encounter this in person. But online, I've seen people say they don't understand Freq method interpretations. As a solution, some suggest to just go with Bayesian methods.
What do you mean by Elementary statistics? The prereq for measure theory requires a solid understanding of proof techniques and Real Analysis.
But a Freq wouldn't say that right? That would be why I think the above is just a strawman of a property they didn't announce they had.
The mere fact that they randomized gives the investigator a good argument for it. Would you agree?
By null results, do they just mean failed to reject? Or do they actually mean the point estimate was 0?
