Re: claims of bimodal distrubutions and the geek gene
Hi All I tried to reply before but I wasn't a member of the list so apparently it didn't get through. Thanks to Chris Douce for adding me, so this reply should work. On 28/11/2009, at 7:43 AM, Raymond Lister wrote: On Fri, 27 Nov 2009, Richard O'Keefe wrote: Anthony Robins http://www.cs.otago.ac.nz/department/staff.php?name=Anthony%20Robins recently noticed http://www.cs.otago.ac.nz/staffpriv/anthony/publications/pdfs/RobinsLEM.pdf Thanks to Richard for mentioning my paper on this. But there are some minor revisions to be made before it is published, so please don't quote the version on my web page quite yet. I'll put the final version up next week some time. And G'day Ray, long time no see! Third, I regard Anthony's model as a good counter argument to the conventional "geek gene" argument, but if I ignore the "geek gene" argument and merely examine Anthony's model in isolation, then I suspect that the number of free parameters in his model means that you can get any distribution you want, by a suitable choice of parameter values. I don't think it has an overabundance of free parameters. Given some pretty minimal assumptions that generate a normal distribution (M0), it requires exactly one additional parameter (an offset, M1) to turn the normal distribution into an anti-normal / bimodal one. This basic model can only vary the "flatness" of the distribution on a continuum from sharply peaked at the centre to normal to flat to bimodal to sharply peaked at the extremes (and, of less interest from my point of view, it can move the centre of the distribution). After that one further parameter shapes the curve at the top or bottom end, but I still don't think that you can quite get "any distribution you want" out if it! I presented one particular version in the paper, but the actual effect is very robust. As Richard described it: "Roughly speaking: a series of topics where failure at any point makes later failure more likely, success makes later success more likely" - almost any reasonable interpretation and implementation of this concept can turn a normal distribution into an anti-normal / bimodal one. As to what, if anything, the model means in educational terms, I make the case in the paper. But Ray's certainly right that there's plenty more to do - I welcome his suggestions as to where to look next, and all other suggestions too... Cheers Anthony Anthony Robins === Computer Scienceanth...@cs.otago.ac.nz University of Otago ph:+64 3 4798314 Dunedin fax: +64 3 4798529 NEW ZEALAND http://www.cs.otago.ac.nz/staff/anthony.html
Re: claims of bimodal distrubutions and the geek gene
This is somewhat tangiential, but I encountered a similar phenomenon in a study I ran recently, and I'd be interested to hear what people on this list had to say about it. Long story short, I gave my subjects a 4-hour window to complete a task, but found every one either did it within two hours, or failed to complete it at all within four (despite continuing to work, apparently understanding the problem, etc). I can't say the distribution is strictly "bimodal", but they definitely fell into two distinct classes. The paper is at http://ecs.victoria.ac.nz/twiki/pub/Events/PLATEAU/Schedule/plateau09-luff.pdf, and the graph is on page 5 (section 4.4). The "success breeds success" explanation does sound like an intuitive fit. I would be very interested to see what others have to say. (Other comments on the study are, of course, also welcome). Meredydd On Fri, Nov 27, 2009 at 6:43 PM, Raymond Lister wrote: > > > On Fri, 27 Nov 2009, Richard O'Keefe wrote: >> Anthony Robins >> http://www.cs.otago.ac.nz/department/staff.php?name=Anthony%20Robins >> recently noticed >> http://www.cs.otago.ac.nz/staffpriv/anthony/publications/pdfs/RobinsLEM.pdf >> that the characteristic bimodal distribution of outcomes in CS1 >> can be explained by a very simple statistical model. >> (Roughly speaking: a series of topics where failure at any point makes >> later failure more likely, success makes later success more likely.) > > First, I recommend Anthony Robins' paper as a refreshing alternative > voice to the repeated claims that passing a zero pre-requisite > programming course requires an innate talent (the "geek gene). But, as > with all research, there are a issues with his work that require follow > up. > > First, I'd like to see evidence collected, from multiple institutions, > examining the grade distribution in CS1. I think the bimodal > distribution is an urban myth. (Would anyone care to formally define > what they mean by a bimodal grade distribution? That's a rhetorical > question, and no correspondence is required.) > > Second, if we are going to make research claims about grade > distributions, then we need to bring research standards to bear on the > contruction of our research instrument - the exam. Most exams are > idiosyncratic constructions, of dubious validity and reliability. > > Third, I regard Anthony's model as a good counter argument to the > conventional "geek gene" argument, but if I ignore the "geek gene" > argument and merely examine Anthony's model in isolation, then I > suspect that the number of free parameters in his model means that > you can get any distribution you want, by a suitable choice of > parameter values. > > Raymond > > Dr. Raymond Lister > Science Teaching and Learning Fellow, > University of British Columbia, > Department of Computer Science, > Vancouver, Canada >
Re: claims of bimodal distrubutions and the geek gene
On Fri, 27 Nov 2009, Richard O'Keefe wrote: > Anthony Robins > http://www.cs.otago.ac.nz/department/staff.php?name=Anthony%20Robins > recently noticed > http://www.cs.otago.ac.nz/staffpriv/anthony/publications/pdfs/RobinsLEM.pdf > that the characteristic bimodal distribution of outcomes in CS1 > can be explained by a very simple statistical model. > (Roughly speaking: a series of topics where failure at any point makes > later failure more likely, success makes later success more likely.) First, I recommend Anthony Robins' paper as a refreshing alternative voice to the repeated claims that passing a zero pre-requisite programming course requires an innate talent (the "geek gene). But, as with all research, there are a issues with his work that require follow up. First, I'd like to see evidence collected, from multiple institutions, examining the grade distribution in CS1. I think the bimodal distribution is an urban myth. (Would anyone care to formally define what they mean by a bimodal grade distribution? That's a rhetorical question, and no correspondence is required.) Second, if we are going to make research claims about grade distributions, then we need to bring research standards to bear on the contruction of our research instrument - the exam. Most exams are idiosyncratic constructions, of dubious validity and reliability. Third, I regard Anthony's model as a good counter argument to the conventional "geek gene" argument, but if I ignore the "geek gene" argument and merely examine Anthony's model in isolation, then I suspect that the number of free parameters in his model means that you can get any distribution you want, by a suitable choice of parameter values. Raymond Dr. Raymond Lister Science Teaching and Learning Fellow, University of British Columbia, Department of Computer Science, Vancouver, Canada