In responseto the article in the Guardian cited byChris Green,
http://www.theguardian.com/commentisfree/2013/oct/22/google-autocomplete-un-women-ad-discrimination-algorithms Rick Fromanwrote: >Idon't begrudge someone advertising for a good cause to use a current trend >like Google autocomplete to make a dramatic adthat makes a good point but >I don't think it is a sociological IAT (or even aRorschach). I triedputting in the three phrases and checking the first hundred or so items foreach, and I found only a very tiny number of sexist responses. An illustrationof the doubtfulness of the "methodology" is that under "womenshould not drive" was a Daily Mail article deriding a Saudi Sheikh sayingwomen shouldn't drive because it would damage their ovaries. Checking "womenshould stay at home", almost all the items quoted it only to criticise it,or discuss the issue of whether it is best for women with young children to stay at home, and so on. There's animportant case to be made here, but I'm not happy about dubious methodologiesbeing used to promote a good cause (for me it comes close to "lying fortruth"). As Rickimplies, this is not of course to deny the need for such a UN campaign, thoughI suspect that the kinds of sexism widelyaccepted in different parts of the world varies considerably, and (dare Isay it) varies to some degree in Western and Northern Europe compared to theUnited States. To make amore general (and serious) point in relation to surveys or self-styled studies on a varietyof social issues, I find it depressing that newspapers (and blogs) that publicisethe results almost never check the methodology to see how solid the conclusionsare. And while I'm at it, I wish sub-editors (andoccasionally journalists) would not write "our" and "we" asif virtually all of us have such attitudes, as in the Guardian's headline "Google'sautocomplete spells out our darkest thoughts". Allen Esterson Former lecturer, Science Department Southwark College, London allenester...@compuserve.com http://www.esterson.org -------------------------------------------------- From: rfro...@jbu.edu <rfro...@jbu.edu> Subject: RE: Google autocomplete & psychology Date: Mon, 4 Nov 2013 16:51:59 -0600 I think Google autocomplete is fairlyunreplicable in many cases. I don't doubt that putting "women should"into a Google search in Dubai might turn upthose suggestions (the UN ad agency was in Dubai), but I wasn't able to replicate ithere. I don't begrudge someone advertising for a goodcause to use a current trend like Google autocomplete to make a dramatic adthat makes a good point but I don't think it is a sociological IAT (or even aRorschach). When I put in "women", mine says: women seeking men women of faith women living well women of the bible When you put in "women should", itstarts to go negative. And what would you expect? What sentence that starts outsaying that a whole class of humanity "should" or"shouldn't" do anything is going to end well? There was an interesting non-disclaimer disclaimerin the Guardian article: "The world as seen through the lens ofautocomplete is a weird and not always wonderful place. It's a place whereDavid Cameron "is a lizard", Obama is "a Muslim", Putin isa "badass" and Miley Cyrus, predictably, is "still twerking".But despite the suggestions that have been skewed by a popular blogpost or memeand are clearly bonkers, there is still much to be gleaned from them about ourdeep-seated discriminations." Really? Much to be gleaned? I doubt it. For "psychology is", I got psychology is psychology is defined as psychology is the study of psychology is best defined as psychology is not a science Certainly there are some people who don'tbelieve it is a science but I don't think 1/4 autocomplete recommendations isvery disheartening. I would think someone was tampering with it if there wasnothing negative at all. Rick Dr. Rick Froman, Chair Division of Humanities and Social Sciences Professor of Psychology Box 3519 John Brown University 2000 W. University Siloam Springs, AR 72761 rfro...@jbu.edu (479) 524-7295 http://bit.ly/DrFroman From: Christopher Green <chri...@yorku.ca> Subject: Google autocomplete & psychology Date: Mon, 4 Nov 2013 17:04:53 -0500 You may have seen, recently, that it has become a kind of informal method for discovering popular social trends to enter the beginning of a sentence into a Google search box and see how the engine completes the expression. The idea is that Google will come up with the most likely completions based on its "knowledge" of what people generally intend when they start a search expression. This method was used very effectively in an ad campaign by UN Women a few weeks ago when they entered phrases like "women should" and "women need to" and got horrible completions like "stay at home," "be slaves," and "be in the kitchen." Here's a Guardian article about that experiment. http://www.theguardian.com/commentisfree/2013/oct/22/google-autocomplete-un-women-ad-discrimination-algorithms So, I decided to try the same thing with "psychology is". You should give it whirl. I don't think you'll find the results to be all that surprising, but they are a bit disheartening nevertheless. Regards, Chris ....... Christopher D Green Department of Psychology --- You are currently subscribed to tips as: arch...@jab.org. To unsubscribe click here: http://fsulist.frostburg.edu/u?id=13090.68da6e6e5325aa33287ff385b70df5d5&n=T&l=tips&o=29793 or send a blank email to leave-29793-13090.68da6e6e5325aa33287ff385b70df...@fsulist.frostburg.edu
