tag:blogger.com,1999:blog-5558726.post2398071967864314599..comments2024-02-20T17:40:21.618-05:00Comments on jeremy freese's weblog: institutional review boards have no jurisdiction over the deadjeremyhttp://www.blogger.com/profile/12755662766163119607noreply@blogger.comBlogger7125tag:blogger.com,1999:blog-5558726.post-29612976163917746722007-03-23T12:56:00.000-04:002007-03-23T12:56:00.000-04:00I hope to God no one who reads this thinks that I ...I hope to God no one who reads this thinks that I don't believe there is a lot of fudging around the edges (and something even less around the edges) in quantitative social science. It's precisely the reason that I'm such a big advocate of data sharing and code sharing.jeremyhttps://www.blogger.com/profile/12755662766163119607noreply@blogger.comtag:blogger.com,1999:blog-5558726.post-84350717427047869082007-03-23T10:36:00.000-04:002007-03-23T10:36:00.000-04:00Point of clarification: by "fudging" I DO NOT mean...Point of clarification: by "fudging" I DO NOT mean "fabricating". I agree with Ang and the processes she outlined and she and Andrea agree on.shakhahttps://www.blogger.com/profile/06003820482910374816noreply@blogger.comtag:blogger.com,1999:blog-5558726.post-76342559089922208222007-03-22T23:47:00.000-04:002007-03-22T23:47:00.000-04:00Oh yeah -- I was agreeing with you, Ang.-andreaOh yeah -- I was agreeing with you, Ang.<BR/>-andreaAnonymousnoreply@blogger.comtag:blogger.com,1999:blog-5558726.post-42478999329227888902007-03-22T23:38:00.000-04:002007-03-22T23:38:00.000-04:00Fudging of data, seeing what you want to see, sele...<I>Fudging of data, seeing what you want to see, selectively representing data -- problematic, sure, but just as possible with quantitative work</I><BR/><BR/>Right. That's what I said.Anghttps://www.blogger.com/profile/07005514932002735616noreply@blogger.comtag:blogger.com,1999:blog-5558726.post-52065637046363861342007-03-22T22:52:00.000-04:002007-03-22T22:52:00.000-04:00Fudging of data, seeing what you want to see, sele...Fudging of data, seeing what you want to see, selectively representing data -- problematic, sure, but just as possible with quantitative work (at least/especially if you're the one actually collecting the data). <BR/><BR/>And it's not like there isn't a record (at least in interview research) -- I have handwritten notes, typed transcripts, and recordings. If anyone wanted to replicate my analysis, they could (no, my data isn't publicly available -- that's another issue). Yeah, I could have made it all up, the same way a survey researcher could make it all up.<BR/><BR/>As to anonymity, I had some respondents expressly ask that I use their real names. In those cases, I did -- because to not do so seemed the paternalism Shamus suggests (and yes, it is scarier that way). In the other cases, I did use pseudonyms, and I'm not entirely sure why this is an "unfortunate" need -- I think you can make a case for using real names (especially when the respondents don't care or prefer it), but otherwise, that's not where I'd call for greater accountability.<BR/>--andreaAnonymousnoreply@blogger.comtag:blogger.com,1999:blog-5558726.post-21632247242750730622007-03-22T17:20:00.000-04:002007-03-22T17:20:00.000-04:00I'm with you Jeremy. For example, in Sidewalk Mitc...I'm with you Jeremy. For example, in Sidewalk Mitch Duneier he argues that changing names of subjects is often more about protecting the ethnographer than it is protecting the subjects. I think he's right about this. <BR/><BR/>Also, arguing that subjects can't consent to having their names revealed strikes me at best, as odd. At worst it is a form of infantalizing paternalism (the researcher knows what's best for you). And it makes me think, what exactly is the RESEARCHER trying to hide? <BR/><BR/>Keep in mind, my ethnographic work is on elites. And no one is very concerned about protecting elites; as the thinking goes: they seem to be doing a pretty good job protecting themselves. I do not protect the name of my site, though I do some of my subjects (as many are workers, they say less than flattering things about bosses, and could be fired). In the writing process I have found that this has made me MORE accountable (and a little scared). But this kind of intellectual accountability demands a degree of honesty that I believe is often missing in social science work. This is most pronounced in ethnographic work where others cannot possibly replicate the study. But it strikes me as being alleviated if ethnographers are more honest about who and what they're studying (provided the subjects consent, and are in a position where they can consent). <BR/><BR/>I am reminded here of Duneier's reply to Wacquant in the great AJS ethnography fight of 2003 when Duneier says (to paraphrase), "If Wacquant is willing to fabricate information from publicly available sources [books] then one wonders how it is possible to even begin to trust what he says about sources that only he has access to". <BR/><BR/>I'm not saying there are NEVER situations where ethnographers should protect their sources by obscuring identity and/or place. But I am saying that I think the practice is far too common. And that the justification of "protecting subjects" can be paternalistic and can be a weak justification for protecting the ethnographer himself. <BR/><BR/>And I think if ethnographers DID reveal their sites/subjects (with consent) they would be taken more seriously, overall, as social scientists. <BR/><BR/>As for fudging data: I think it's widespread.<BR/><BR/>-Shamus Khan (for some reason, blogger won't let me sign in!)Anonymousnoreply@blogger.comtag:blogger.com,1999:blog-5558726.post-40701154973515673212007-03-22T16:22:00.000-04:002007-03-22T16:22:00.000-04:00Interestingly, I've found that suggesting the exis...Interestingly, I've found that suggesting the existence of fudged data or data analysis (notwithstanding the assumed occasional crackpot or evildoer) as being at least somewhat common, is met with either blank stares or defensive statements questioning <I>my</I> integrity. <BR/><BR/>But isn't it actually kind of reasonable to assume that there's at least some decent-but-unspecified amount of fudging going on, in both qual and quant work? Not just because people are desperate for success, but also because humans like to find things they're looking for? Some of it's bound to be inadvertent.<BR/><BR/>Anyway, this is an interesting topic. I'll be interested to hear what the JFW readers think.Anghttps://www.blogger.com/profile/07005514932002735616noreply@blogger.com