vasishth
commented
1 year ago Yes, this was really helpful! Thanks! I have edited what I could right away and opened an issue for material that requires my re-reading the chapter.
Closed omseeth closed 1 year ago
vasishth
commented
1 year ago Yes, this was really helpful! Thanks! I have edited what I could right away and opened an issue for material that requires my re-reading the chapter.
Here are a few typos I spotted in chapter 2.
Section 2
[1] "a crucial point: the posterior…“ -> "the“ should be capitalized
Section 2.2.5
[2] "Visualizing the prior, likelihood, and the posterior“ -> should probably be „the prior, likelihood, and posterior“
Section 2.5
[3] "distributed data points from the Geometric distribution, which has the likelihood function“ -> missing colon behind "function“.
[4] "data points x can have values 0,1,2,.“ -> sentence ends with ",.“
I also have the following thoughts regarding this chapter.
Section 2.2
[1] It could be clarified that the following statement: "The repeated runs of the (simulated) experiment are the sole underlying cause for the variability in the estimated proportion“ relates to the "sd(estimated_means)“ given in the example. Maybe adding something like "as is visible in the standard deviation“ would suffice.
[2] Maybe that’s the philosopher in me speaking but when you say "to represent our prior belief or knowledge about plausible values“ I’m wondering: Is it a belief or is it knowledge? For these two concepts traditionally have different meanings. -- Or under what conditions is the prior a belief and when knowledge?
Section 2.2.3
[3] I’m not sure whether it’s on purpose, but I’ve noticed a varying use of the multiplication sign: Sometimes it is a dot, sometimes an x. For example, (2.6) appears to make use of both signs in the same equation.
[4] On this note, I’ve also noticed a varying use of parameters in R code. For example in chapter 1 in section 1.4, you write "rbinom(10, n = 20, prob = 0.5)“. In section 1.4.2.4, we have "rbinom(1, size = 10, prob = 0.5)“.
Section 2.2.6
[5] I think it could be a good idea to remind the reader that increasing the parameters a and b is what would fall under prior specification right at the beginning of section 2.2.6. Because a reader new to the topic could have forgotten at this point what you had mentioned in 2.2.2. Then he would wonder by the end of 2.2.6. what you mean by "Whenever we do a Bayesian analysis, it is good practice to check whether the parameter you are interested in estimating is sensitive to the prior specification.“ and would have to search the text again for the explanation of the concept.
I hope this helps!
Best, Max