Correlation & causation

[Elaedith]

Does anyone know some good examples of correlation being reported as causation in the media, that can be found online. I am looking for examples that would be interesting to social science students, for a seminar on 'critical thinking about correlation'.

[Jeff Corey]

These have come in handy, http://jonathan.mueller.faculty.noctrl.edu/100/correlation_or_causation.htm

[Modified]

Does anyone know some good examples of correlation being reported as causation in the media, that can be found online. I am looking for examples that would be interesting to social science students, for a seminar on 'critical thinking about correlation'.

Health news would be your easiest target. Search Google news for: health news correlation causation , and you'll find numerous articles on the problem. Sometimes the culprits are the study authors, but it's usually the news article writers.

[bpesta22]

I like these:

spanking and IQ
nations adding flouride to water and cancer rates.
In florida, couples who stay married this year are more likely to die (Marriage kills !).

[bpesta22]

Also, any sort of home remedy that works because I did it and my zit cleared up two days later (or cold went away two days later).

My all time favorite is the use of morning urine on zits.

[athon]

Probably the best recent example is the death of UK school girl, Natalie Morton (http://www.dailymail.co.uk/news/article-1216714/Schoolgirl-14-dies-given-cervical-cancer-jab.html), hours after getting a Gardasil injection. It was found that she had suffered from an extremely aggressive tumour, however that didn't stop newspapers (and antivaxers) from leaping to conclusions.

Athon

[arthwollipot]

This was exactly what I was about to post, athon. There was another one recently that we covered in the latest episode of the podcast (siglinked) about a girl who developed dystonia after a flu injection.

[Jorghnassen]

Probably the best recent example is the death of UK school girl, Natalie Morton (http://www.dailymail.co.uk/news/article-1216714/Schoolgirl-14-dies-given-cervical-cancer-jab.html), hours after getting a Gardasil injection. It was found that she had suffered from an extremely aggressive tumour, however that didn't stop newspapers (and antivaxers) from leaping to conclusions.

Athon

Except that's not a correlation (which implies statistical evidence), it's just an anecdote, a coincidence attached to a post hoc ergo propter hoc (while the correlation causation fallacy is cum hoc ergo propter hoc).

But here's a closer example (http://www.forbes.com/2003/06/18/cx_mh_0619oscar.html): winning an Oscar adds 3.9 years to one's life. When you forget about immortal time bias... Mind you, the original authors made the mistake, not the media in particular.

[six7s]

Quid quid latine dictum sit, altum videtur

[A.A. Alfie]

You mean

Quidquid latine dictum sit, altum videtur

[athon]

Except that's not a correlation (which implies statistical evidence), it's just an anecdote, a coincidence attached to a post hoc ergo propter hoc (while the correlation causation fallacy is cum hoc ergo propter hoc).

While you're not wrong, I think it'd work perfectly for the OP's needs. It is still technically a correlation, even if it's not based on a relatively large data pool. And it's quite accessible to most people, perfect for sparking a discussion amongst students.

Sometimes I think skeptics get too caught up in latin phrases and trivial technicalities, missing opportunities to engage with people who don't have the experience in thinking critically.

Athon

[six7s]

http://imgs.xkcd.com/comics/correlation.png
xkcd.com/552/ (http://xkcd.com/552/)

[bpesta22]

While you're not wrong, I think it'd work perfectly for the OP's needs. It is still technically a correlation, even if it's not based on a relatively large data pool. And it's quite accessible to most people, perfect for sparking a discussion amongst students.

Sometimes I think skeptics get too caught up in latin phrases and trivial technicalities, missing opportunities to engage with people who don't have the experience in thinking critically.

Athon

Technically, you mean it's a co-variance as neither variable was standardized.

:duck:

[athon]

Technically, you mean it's a co-variance as neither variable was standardized.

:duck:

:p

I rest my case, your honour.

Seriously, though, it does bring up again the subtle variation between colloquial understanding of a term and the role of jargon. Which is, essentially, what my world consists of currently - I'm always having to take a word as it's used in a technical sense and translate it, or vice versa.

Meh, anyway...sorry for the derail.

Athon

[cbish]

As an educator, sadly I must say, most educational research seems to fall under this category.

[aggle-rithm]

What about the article about the scientists who taught a grasshopper to jump when they clapped, then pulled off the grasshhopper's legs, then when it failed to jump concluded that legless grasshoppers are deaf?

That was an article, right?

[Elaedith]

Thanks for the suggestions. I was thinking of statistical correlation but we do also relate the issue to reasoning errors like post hoc fallacy.

[athon]

Here's an interesting one I just ran across.

http://www.good.is/post/rock-and-u-s-oil-production-is-dead/

Athon

[justcharlie09]

What about the article about the scientists who taught a grasshopper to jump when they clapped, then pulled off the grasshhopper's legs, then when it failed to jump concluded that legless grasshoppers are deaf?

That was an article, right?

LOL... and to the OP...as a former psych major who plans to do ANYTHING but psych as grad work...you have no idea how much I'm enjoying this thread. Thank you.

[Jeff Corey]

What about the article about the scientists who taught a grasshopper to jump when they clapped, then pulled off the grasshhopper's legs, then when it failed to jump concluded that legless grasshoppers are deaf?

That was an article, right?

No. That was an old joke, usually with the ethnicity of the scientist la gout de jour..