tag:blogger.com,1999:blog-9263167.post6245264236649671946..comments2024-03-28T15:17:43.056-04:00Comments on Stayin' Alive: Here's some boring wonkery . . .Cervanteshttp://www.blogger.com/profile/11302076828795198187noreply@blogger.comBlogger3125tag:blogger.com,1999:blog-9263167.post-86377605286774587392011-04-05T10:02:17.782-04:002011-04-05T10:02:17.782-04:00Well now David, you don't give enough of the p...Well now David, you don't give enough of the particulars for me to interpret your story. (I am quite familiar with principal components analysis, BTW.)<br /><br />You lied to the man. Apparently, by coincidence, your random data ended up looking like something interpretable. It's not the professor's fault, you gave him particular results. You also don't say how the real data turned out different, or whether you explained your deception. <br /><br />Why are you skeptical of the scientist when it is you who committed the fraud? I don't really get this.Cervanteshttps://www.blogger.com/profile/11302076828795198187noreply@blogger.comtag:blogger.com,1999:blog-9263167.post-16082303402397989092011-04-04T11:51:09.658-04:002011-04-04T11:51:09.658-04:00i am too jaded to even pretend to be shocked at th...i am too jaded to even pretend to be shocked at the dishonesty of big business.<br /><br />both my mother and mother-in-law have suffered discomfort and dangerous medical conditions from (foolishly? uninformed?) prescriptions. and we have had to intervene to get the "treatment' to stop.rogerhttps://www.blogger.com/profile/14554961854201609452noreply@blogger.comtag:blogger.com,1999:blog-9263167.post-49808588425929312302011-04-03T05:32:53.817-04:002011-04-03T05:32:53.817-04:00Some while ago, I wrote a program for Factor Analy...Some while ago, I wrote a program for Factor Analysis that was of considerable interest to the "social scientists" who collected vast troves of data, confident that contained immutable truth if only they knew how to ask the right questions of it. Part of the process is to reduce the data to a correlation matrix. On one occassion, just for fun, I substituted random numbers (<1.0) on the off diagonal elements of this matrix. The academic concerned was very excited about the results, as they confirmed his intuitions about the nature of the process he was studying. As he was intending to present the results to a conference in a week, I quickly reran the program with the original data. No change in the enthusiasm; no change in the paper presented. The only change was that I became very sceptical about the claims of so called "soft sciences"DavidKNZhttps://www.blogger.com/profile/00795616145927146882noreply@blogger.com