Yeah he thought if you composed a bunch of criminals photos, you'd be able to identify the facial features correlated with criminality (and thus, I suppose from a eugenicist perspective, remove them from the gene pool)
Anyway, turns out that the more images you compose, the better looking people tend to become, so that's how first learned that what people find beautiful tends to be an average of people's faces. posted by delmoi at 9:25 AM on April 25, 2012
Similarly for professions—what does The Barber look like?
It works way better on families, since you get something like either: (a) a plausible candidate for That Jones Look, you know, the thing they all have in common, or (b) a plausible candidate for just another member of the Jones family. posted by kenko at 9:29 AM on April 25, 2012
Because I am a pedant: Pearson, not Galton, invented the correlation coefficient in the sense of coming up with the formula for it. Galton seems to have had some idea of some sort of coefficient measuring the strength of the regression to the mean in his 1886 paper on heights of parents and children, though. (Disclaimer: I have only skimmed the paper I just linked to.)
I can't figure out how to get a linkable page, so let me just encourage everyone to search "Punch" and go to the sixth result, and you will get my favorite thing Galton ever wrote. posted by Sidhedevil at 9:32 AM on April 25, 2012
If I understand correctly (and it's good for me to understand this correctly, because I'm writing about Galton right now, which is why I was on this page) Galton defined correlation coefficient for two variables which were assumed to be distributed normally, and Pearson understood that it could be defined purely on the observed values without any prior assumption on the distribution.
The article madcaptenor links to is awesome and clarified the history for me greatly.
"By now, a century later, contemporary scientists often take the correlation coefficient for granted. It is not appreciated that before Galton and Pearson, the only means for
establishing a relationship between variables was to educe a causative connection. There was no way to discuss-let alone measure-the association between variables that lacked a cause-effect relationship." posted by escabeche at 9:37 AM on April 25, 2012
You know, I need to post Francis Galton and his Beauty Map. posted by jadepearl at 10:26 AM on April 25, 2012
"Inventor of fingerprint identification" - I would credit Hershel with that. Galton just classified it. posted by unliteral at 5:44 PM on April 25, 2012
Galton was a creative scientist, but his reputation as the grand-daddy of Eugenics casts a shadow upon his legacy... posted by ovvl at 7:43 PM on April 25, 2012
Mark Twain wrote a story with fingerprint identification of a murderer even before Sir Francis Galton published his studies.
In Natural Inheritance (1889) he presents the regression-toward-the-mean theory: "...it is theoretically a necessary fact, and one that is clearly confirmed by observation, that the Stature of the adult offspring must on the whole, be more mediocre than the stature of their Parents; that is to say, more near the M of the general Population."
So I guess saying you are descended from someone like Sir Francis Galton doesn't really say much. posted by eye of newt at 10:50 PM on April 26, 2012
« Older Harvard sociobiologist E. O. Wilson explores The O... | Hopefully the new Avengers fil... Newer »
This thread has been archived and is closed to new comments
Composite photography!
posted by kenko at 9:00 AM on April 25, 2012 [1 favorite]