Skepticism about beautiful people having more daughters. | Hacker News Reader

Skepticism about beautiful people having more daughters.

stat.columbia.edu

43 pointslliiffee9 comments

It's great to have what effectively amounts to a rebuttal on the front page at the same time as the article it disputes.

I find it great that on HN, new links are often relevant to recently, previously discussed links. You can almost see the collective mind digesting new information and putting out new conjectures.

Still, I wonder if there would be a good system for showing two stories are related, because although I frequently check HN, I don't always keep up and some stories really need the context of the previous story.

I wonder if there would be a good system for showing two stories are related

I believe that is called a hyperlink, as here.

http://news.ycombinator.com/item?id=726421

Hyperlinks generalize to showing that three or more stories are related, and allow annotations (by accompanying text) to show the nature of the relationship. Here, I can make clear that the thread to which I post this reply links out to an article that refutes the article linked out to in the link

http://news.ycombinator.com/item?id=726421

I have just provided to another HN discussion thread.

Hypertext is wonderful.

All right, I see the downvotes, so let me ask a question: what other technical means do people have in mind? Today's existing technology allows any of us to show relatedness of any URL to any other URL by posting the link and accompanying that link with explanatory text. What other technology would require less human thought about how to show that two (or more) recent threads here on HN are related?

As an example, could I ask for a demonstration of relatedness between my post above and any part of the HN guidelines?

You were a bit sarcastic and cheeky, which doesn't play well, but your point is well taken so I have upvoted you a bit.

Perhaps the submission form could use a "related stories" field if you're submitting a response to a previous story or something?

Perhaps the submission form could use a "related stories" field if you're submitting a response to a previous story or something?

Thanks for the specific suggestion. Along those lines, on this and other forums I've often wondered how technically feasible it is to mine the database as a submission is received, so that the user submitting a new article obtains a notification like "This submission looks very much like . . . " showing links to previously submitted articles that by titles (or keywords in full text) seem to be about the same subject. That would allow the user to

a) decline to submit the new link altogether,

b) indicate which previous links are related (as you suggest)

c) merge into one or another existing thread,

d) perhaps some other action that doesn't immediately come to mind.

But that still requires human thought by the programmer of the forum, at least, and most likely by the submitter as well. Sorry about the snarkiness above, but that's where the snarkiness came from--technical solutions fail if users don't have the commitment to make use of them.

This letter:

http://www.stat.columbia.edu/~gelman/research/published/kana...

effectively shows that the original paper ignored endogeneity, mis-interpreted coefficient estimates and data-mined.

I think most people try a variety of things in their life, eventually hit on something that yields good returns, and then milk it for all it's worth. So that "beautiful daughters" guy has found a gold mine. Who cares about science - tell people what they want to hear, and what they can use to chat on parties, and you can make lots of money. (Not only selling books, but also getting grants and invitations to universities, no doubt).

With that thesis the author can get famous, where is his incentive to be honest about it?

I am waiting for the corresponding articles to appear in all major magazines in my country (presented as established fact, of course). It just makes for too good a story to pass on it.

With that thesis the author can get famous, where is his incentive to be honest about it?

Well, this article is about a scientist, or statistician, alleging that the results are "speculation". She is being polite of course and in the scientific lingo she is probably saying that he is a fraud. So that is his incentive, he can be honest and find a "true" goldmine, or gamble and either go to the top or sink to the bottom.

Reply on news.ycombinator.com