[FoRK] the plural of anecdote is data
baisley at alumni.rice.edu
Tue May 15 00:05:26 PDT 2012
On May 10, 2012, at 6:18 AM, Dave Long <dave.long at bluewin.ch> wrote:
> Fig 1. on the 3rd page of "Do Values Grow on Trees?" shows relatively different editing behaviors for (a tiny selection of) students who went on to receive, respectively good, fair, and poor grades in a programming class.
> The interesting thing is that (consistent with industry practice) the good student spent a much larger fraction of time with intermediate programs that, whether or not they had the desired functionality, at least compiled.
> My speculation is that programmers may be like riders, in that a classical goal of riding is to be in such an equilibrium that an outside observer does not notice the dialog between horse and rider (software is like sausage: the people who consume it often have little to no interest in how it is produced), and the great distinction between good and poor riders is that when a horse requires correction, a strong rider quickly restores the invariant that the horse is at least doing *something* in good form (even if it remains to patiently, incrementally, teach what was desired in the first place), while a weak rider expends a great deal of time in which the horse is not doing anything in particular besides being confused about what might possibly be expected of it.
> Do good programmers make frequent, effective even if occasionally not subtle, interventions in their programs, while poor programmers make infrequent, ineffective ones?
>  Marceau, Fisler, Krishnamurthi "Do Values Grow on Trees? Expression Integrity in Functional Programming" 2011
I would like to include this in my Job-Seekers' Compendium, "Do References Grow On Trees?". I'm also contemplating a Social-Throw-Weight Companion, "Do Names Reduce In Maps?", which may/not include a deconstruction of the 1972 Melanie hit song, "Brand New Key:Value Pair".
More information about the FoRK