tag:blogger.com,1999:blog-35248477.post5177097651988456290..comments2024-03-27T14:20:05.905-04:00Comments on Montclair SocioBlog: The AssociationJay Livingstonhttp://www.blogger.com/profile/06652075579940313964noreply@blogger.comBlogger3125tag:blogger.com,1999:blog-35248477.post-26293815196606180012009-02-26T17:11:00.000-05:002009-02-26T17:11:00.000-05:00Re: "WVU Prof Linked to Organized Crime"Damn! I'v...Re: <I>"WVU Prof Linked to Organized Crime"</I><BR/><BR/>Damn! I've been discovered. I'm with you Jay. I'm waiting for the article to come out so I can examine the argument in more detail. Off the cuff, I'm willing to buy the "linkage" between lyrics and behavior, insofar as lyrics are a proxy for consumption behavior of certain groups of people. But if I'm right, then the linkage isn't between the lyrics... (Dr. Dre made me do it); it's between the social groups and the behavior.Coreyhttps://www.blogger.com/profile/00119612449610307692noreply@blogger.comtag:blogger.com,1999:blog-35248477.post-85272272899402951682009-02-26T13:46:00.000-05:002009-02-26T13:46:00.000-05:00Corey, Here's the link to the post about the artic...Corey, Here's the link to the post about the article http://www.physorg.com/news154683399.html. (The article itself is not yet available at the journal Website.) The title of the post is "Sexual lyrics in popular songs linked to early sexual experiences." "Linked" is one of those words you here when the speaker wants to imply a stronger connection than is probably there ("WVU Prof Linked to Organized Crime").Jay Livingstonhttps://www.blogger.com/profile/06652075579940313964noreply@blogger.comtag:blogger.com,1999:blog-35248477.post-27294078700472868122009-02-26T13:38:00.000-05:002009-02-26T13:38:00.000-05:00Re: "The emphasis is my own addition because someb...Re: <I>"The emphasis is my own addition because somebody here is missing a point that any intro sociology student should have learned: correlation is not cause.</I>.<BR/><BR/>[Sarcasm]<BR/>Yeah, but Jay... the correlation is significant at .05 level... that means that it's independently true of any criticism... right? <BR/>[/sarcasm]<BR/><BR/>Sigh... actually correlation should be used to establish evidence of causation. At least that's the way it's supposed to work, when the model has been properly specified on a well established theoretical foundation. I haven't read the article in question that establishes "independent correlation", but I would hope that the authors actually did some conceptual/theoretical work in establishing that theoretical linkage between rap and sex. My guess is that they didn't, or did so in such a thin manner that it is the equivalent of not theorizing at all.<BR/><BR/>My problem with much of modern "empirical" (and I use that term loosely) social science is that it skips the whole conceptualizing process. Step 1: Get a topic; Step 2: Get a data set; Step 3 run some math on the data set; Step 4, create hypotheses post-hoc that fit the statistically significant associations; Step 5 write it up as if the hypothesis was established on the front end.<BR/><BR/>Incidentally, Jeremy Freese wrote a very readable overview of good practice for survey analysis in his chapter on the Secondary Analysis of Large Survey Data Sets.Coreyhttps://www.blogger.com/profile/00119612449610307692noreply@blogger.com