On 27 Mar 2003 at 13:36, Dennis Roberts wrote: > At 03:01 PM 3/27/03 -0500, Donald Burrill wrote: > > (in more detail of course) ... that methods AND results are both important > and i fully agree > > the purpose of courses and training along these lines is to move up the > scale on both ... that is, improve your methods for solving problems AND > improving your odds of getting the correct solutions > > trying to separate out each side of this method/solution equation seems > like folly to me > > which brings me to the following > > this is a big issue with the USE of software ... it is so easy to produce > reams of output ... but, have no idea what it is .... nor how it was derived > > while i am a big fan of software ... i also recognize the downside ... > because, too often ... instructors say: "well, if you need a t test, here > is the button you press in software package A" ... it is almost TOO easy > > now, i don't want to go back to the days when all this work has to be done > by hand (well, it can't anyway) but, i do think we have to forgo the quick > movement of the students TO the software before we have made sure they have > some basic understanding of what is being done
Use of computers most places is I think firstly a resource question. Given you hace enough computers, there is now software which can actually HELP in understanding the concepts. Simulations is a powerful tool for this, like simulating confidence intervals. Also it makes it possible to introduce data analysis concepts earlier, like residual analysis, and the excellent (and free) program VISTA (www.visualstats.org) has what is named guidemaps, graphics to guide the student throug an analysis. This should make it possible to introduce concepts about strategy of data analysis earlier. The problem is probably that computers is used to teach the same content as pre-computer age, then one cannot expect much. Kjetil Halvorsen > > of course, was that not the debate that raged over calculators? > > . > . > ================================================================= > Instructions for joining and leaving this list, remarks about the > problem of INAPPROPRIATE MESSAGES, and archives are available at: > . http://jse.stat.ncsu.edu/ . > ================================================================= . . ================================================================= Instructions for joining and leaving this list, remarks about the problem of INAPPROPRIATE MESSAGES, and archives are available at: . http://jse.stat.ncsu.edu/ . =================================================================
