On Wed, Jun 23, 2010 at 8:59 PM, Geoffrey Smith <[email protected]> wrote: > You might think about looking at stock returns and asking questions about > the probability of doubling your money or of getting a rate of return that > exceeds the historical mean. Geoff
Doesn't always work though. Bio-engineers for example are typically less interested in gambling on the stockmarket... ;-) Now serious, one of the finest examples I met is the example about the death penalty in Florida, to be found in Agresti's book "Categorical Data Analysis" on page 48 and further. It gives a very nice illustration of Simpson's paradox, applied to real data about a topic that leaves few untouched. Basically Agresti shows there how one came to the conclusion that whites got more often the death penalty than blacks in Florida. If one didn't take the victim into account, that is. When looking at the victims, it was clear that the death penalty was more often declared when the victim was white. It turned out that blacks mostly kill blacks, and whites mostly killed whites. Conditional on the victim, the story turned around and blacks clearly got the death penalty more often. The data goes from 1971 to 1987, so is fairly safe to use for educational purposes Cheers Joris > > On Wed, Jun 23, 2010 at 11:09 AM, Christophe Genolini <[email protected] >> wrote: > >> Hi the list, >> >> As a statistics teacher, I teach to NOT-scientists student, public that it >> is permanently necessary to motivate. I am thus in search of examples both >> scientific and playful to illustrate my courses. It is not always easy to >> find. As other teacher might be in the same case, I say to myself that we >> could maybe share our 'best' examples? >> >> So I start: a social psychologist (Nicolas Gueguen, article here >> http://nicolas.gueguen.free.fr/index.html) has establishes that if we >> approach a perfect unknown lady on a beach and we ask for its phone number, >> we have 9 % of chance to obtain it. If we call her by touching her slightly >> on the front arm, we have 19 % (!!!) of chances to obtain it (test of chi2, >> p < 0.01). Surprising, isn't it? >> >> So what are your 'best examples' ? >> >> Christophe Genolini >> >> -- >> ----------------------------------------- >> Christophe Genolini >> Maitre de conférences >> INSERM U669, Equipe Biostatistiques >> UFR STAPS, Université de Paris Ouest-Nanterre-La Défense >> Web: http:\\christophe.genolini.free.fr >> >> _______________________________________________ >> [email protected] mailing list >> https://stat.ethz.ch/mailman/listinfo/r-sig-teaching >> > > > > -- > Geoffrey Smith > Visiting Assistant Professor > Department of Finance > W. P. Carey School of Business > Arizona State University > > [[alternative HTML version deleted]] > > > _______________________________________________ > [email protected] mailing list > https://stat.ethz.ch/mailman/listinfo/r-sig-teaching > > -- Joris Meys Statistical consultant Ghent University Faculty of Bioscience Engineering Department of Applied mathematics, biometrics and process control tel : +32 9 264 59 87 [email protected] ------------------------------- Disclaimer : http://helpdesk.ugent.be/e-maildisclaimer.php _______________________________________________ [email protected] mailing list https://stat.ethz.ch/mailman/listinfo/r-sig-teaching
