« first day (3077 days earlier)      last day (1938 days later) » 

1:18 AM
What do you have to say about this statement: "Any hypothesis that withstands any test of refutation becomes more likely"?
 
2:12 AM
One of the posts on the HNQ is currently "Why are aircraft parts built in different places and assembled in one?" ....
I want to answer "because if you assembled them in two places you'd have two half-planes".
 
 
10 hours later…
12:07 PM
Can you please downvote the following answer? stats.stackexchange.com/a/244021/82135. I am more confused than before after having read it.
 
 
4 hours later…
4:18 PM
@nbro With respect, that you might be confused by an answer is insufficient basis to encourage the entire community to downvote it! This particular answer creatively addresses the question by providing a single clear example illustrating the "difference between prediction and inference." I would encourage people to upvote this answer (which I did long ago).
2
 
4:40 PM
@whuber Do you think that providing multiple examples or answers will help in clarifying the doubts that a person might have because of the numerous existing and often not consistent definitions?
For example, what is a measurement? How is it different from a dataset? This answer is far from being clear, IMHO.
Furthermore, defining inference in terms of "infer" is like a self-reference and usually does not help.
Or, at least, define "infer" first or say that you took the infer definition from the dictionary.
 
5:11 PM
@nbro I believe that different people learn things differently and have different experiences and levels of knowledge. Accordingly, what works for one might not work for another. That, however, is no reason to impugn an answer altogether. It is unheard of in our community for anyone to encourage downvotes or even to suggest downvotes are needed unless an answer is flagrantly wrong.
 
@whuber Ok, maybe I should not have encouraged you to downvote. Maybe I should have just asked why that answer has been upvoted, given that, IMHO, it is unclear (because of what I just said).
Note that no answer was accepted, and the question was asked 2 years ago. Maybe you should wonder why the OP did not originally accept any answer. I guess he was not satisfied with any of the answers given (and there are many). Anyway, answers targeting people that are already quite familiar with the concepts are quite useless.
 
 
1 hour later…
6:34 PM
@nbro One shouldn't read too much into lack of acceptance when this is the sole question the OP has asked. In many (if not most) cases that's due to unfamiliarity with the site.
2
 
 
2 hours later…
8:29 PM
It would be nice if someone can give an answer to this question.
0
Q: Why do we care about the joint distribution of the endogenous variables of a causal model?

nbroIn general, we can calculate the joint distribution of the endogenous variables of a structural casual model (SCM) as follows $$ P(X, X_1, \dots, X_n) = \prod_i = P(X_i \mid \text{parent}(X_i)) $$ where $\text{parent}(X_i)$ is the parent node of $X_i$ in the graphical representation of the SCM...

 

« first day (3077 days earlier)      last day (1938 days later) »