RonM wrote: > More off topic. . . > > I'd have thought that would be the case only if there was some > systematic non-randomness in the original assignment of rightness to B > or C. IF that assignment is truly random, then it doesn't matter what > mechanism the test taker uses to determine which to choose. > > If the original assignment was done on some non-random basis, problems > of course arise, but only if all professors tend to choose the same > method, e.g. never make C the right answer. If some always use C and > some always use B, and this proportion is roughly proportional or can't > be known, it still shouldn't make any difference what method the test > taker uses to determine the selected option if what's true isn't > actually known. > > R.
Hmmm, so a Bernoulli trial where one could calculate the overall probability of guessing independently on each question in an 'N' item exam where the probability of a correct pure guess being .25 for each trial (4 options) as compared to guessing "C" on all N items. But then I guess it gets messy because I'd have to make some assumptions about the distribution of "C" as correct answer across the the N items (which is your point about what professors tend to choose). I need more coffee. ;-) ------------------------------------------------------------------------ garym's Profile: http://forums.slimdevices.com/member.php?userid=17325 View this thread: http://forums.slimdevices.com/showthread.php?t=98010 _______________________________________________ audiophiles mailing list audiophiles@lists.slimdevices.com http://lists.slimdevices.com/mailman/listinfo/audiophiles