Timeline for Does correlation = 0.2 mean that there is an association "in only 1 in 5 people"?
Current License: CC BY-SA 3.0
32 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Feb 18, 2018 at 15:32 | comment | added | Firebug | People, people are fallible. This air of superiority isn't helpful at all. | |
| Feb 17, 2018 at 22:28 | history | edited | amoeba | CC BY-SA 3.0 |
edited title
|
| Feb 17, 2018 at 14:41 | comment | added | PGnome | @DavidRicherby shrug I would hope so, but I’ve seen more egregious misunderstandings before. Quotes are notorious for being taken out of context, I figure it doesn’t hurt to confirm. Comments are for clarifications, no? | |
| Feb 17, 2018 at 14:23 | comment | added | David Richerby | @pwcnorthrop If the author immediately explained why the claim is wrong, I doubt the question would have been posted. | |
| Feb 16, 2018 at 23:31 | comment | added | Russ Lenth | I think somebody should write to that guy and tell him that, as a service to science, he should stop “explaining” statistical concepts. | |
| Feb 16, 2018 at 22:27 | comment | added | Bryan Krause | I'm a neuroscientist. Now I'm an embarrassed neuroscientist. | |
| Feb 16, 2018 at 22:18 | answer | added | Acccumulation | timeline score: 2 | |
| Feb 16, 2018 at 20:36 | comment | added | PhD | I'd stop reading that book, right about now | |
| Feb 16, 2018 at 14:21 | comment | added | PGnome | Is there any context to this quote? The " associated in only 1 in 5 people", is, of course, nonsense, but could it be the author is repeating a misunderstanding he has heard before and then explains what is wrong? | |
| S Feb 16, 2018 at 13:59 | history | suggested | Rodrigo de Azevedo | CC BY-SA 3.0 |
Minor edits
|
| Feb 16, 2018 at 12:57 | review | Suggested edits | |||
| S Feb 16, 2018 at 13:59 | |||||
| Feb 16, 2018 at 10:50 | comment | added | Iwillnotexist Idonotexist | @whuber There are many other ways as well - when I saw this question my first instinct was to provide a proof by classic counterexample. When $x \sim \textrm{Uniform}(-1, 1)$, $x$ has 0 correlation with $y = x^2$, yet clearly there is a direct "association" between $x$ and $y$ in all cases. It's also a nice way to segue into the difference between independence and (un)correlation. | |
| Feb 15, 2018 at 22:32 | comment | added | whuber♦ | @amoeba Either one, depending on circumstances. Or even the open-ended approach of giving the quotation and asking for a comment on it. | |
| Feb 15, 2018 at 22:19 | comment | added | amoeba | @whuber Sounds like a good idea but how exactly would you turn this into an interview question? Do you mean asking "What does rho=0.2 mean?" or do you mean asking, as the OP here, "A book says that rho=0.2 means association in 20% of people, is that correct?" | |
| Feb 15, 2018 at 21:49 | comment | added | Meni Rosenfeld | @MattKrause: This reminds me of the (topical for this question!) saying "Some people believe we only use 10% of our brain. This might be true for the people believing it." | |
| Feb 15, 2018 at 16:25 | comment | added | whuber♦ | I have favorited this post because it's precisely the kind of extremely simple question that, when asked of a stats 001 student (or any other neophyte, or a job applicant), will instantly and unmistakably determine whether they understand what correlation means. | |
| Feb 15, 2018 at 16:07 | comment | added | Matt Krause | @NickCox, to be fair, the book's title does hint that an idiot is somehow involved in writing it :-) | |
| Feb 15, 2018 at 13:41 | vote | accept | Sitak | ||
| Feb 15, 2018 at 13:19 | comment | added | Nick Cox | This 0.01% sample of the book makes me wonder what nonsense is to be found in the rest... | |
| S Feb 15, 2018 at 13:06 | history | edited | gung - Reinstate Monica | CC BY-SA 3.0 |
clarification
|
| S Feb 15, 2018 at 13:06 | history | suggested | smci | CC BY-SA 3.0 |
clarification
|
| Feb 15, 2018 at 12:28 | review | Suggested edits | |||
| S Feb 15, 2018 at 13:06 | |||||
| Feb 15, 2018 at 0:27 | history | tweeted | twitter.com/StackStats/status/963932861203206144 | ||
| Feb 14, 2018 at 21:00 | answer | added | AdamO | timeline score: 9 | |
| Feb 14, 2018 at 20:49 | answer | added | Aksakal | timeline score: 18 | |
| Feb 14, 2018 at 20:39 | comment | added | James Phillips | 4 percent makes much more sense than 20 percent, thank you kindly for the correction, I agree with you. | |
| Feb 14, 2018 at 20:29 | answer | added | Kodiologist | timeline score: 70 | |
| Feb 14, 2018 at 20:28 | comment | added | Richard Hardy | @JamesPhillips, what you are referring to is $r^2$, not $r$ itself. If $r=0.2$ then $r^2=0.04$ so 4%. | |
| Feb 14, 2018 at 20:27 | comment | added | James Phillips | Would it make more sense that 20 percent of the variation in intelligence can be explained by height? | |
| Feb 14, 2018 at 20:23 | history | edited | Firebug | CC BY-SA 3.0 |
added 22 characters in body; edited tags
|
| Feb 14, 2018 at 20:08 | review | First posts | |||
| Feb 14, 2018 at 21:34 | |||||
| Feb 14, 2018 at 20:07 | history | asked | Sitak | CC BY-SA 3.0 |