[Bluej-discuss] Programming as random process ...
Jay Fenwick
jbf at cs.appstate.edu
Thu Apr 20 04:26:40 BST 2006
Gordon, I don't really know the answer to your question. However, I
chuckled because your query reminded me of a moment earlier this semester
with my own CS 1 class (we use BlueJ with Barnes/Kolling text for mostly
the same reasons you stated).
This term I have been giving a 10-12 minute quiz (microAssignment) almost
every class session. Typically they have to write code (very little
terminology type questions). Early on it might just be statements, a
little later it's a simple class with a couple fields and a constructor.
Then I might give them that much and ask for mutator and accessor methods.
They see "whole" classes (albeit early versions) ALOT. Questions
are repeated on subsequent quizzes until all (or nearly) get it correct.
Okay, now the story. One student complained that the quizzes were unfair.
When I asked why he said that in the quiz he didn't have BlueJ there to
guide him to a correct solution. He said he did EXACTLY what you
complained about - random changes to make BlueJ happy. After I (and the
rest of the class) stopped laughing, I explained that he should use
BlueJ's compiler to guide his efforts; but not random changes, rathar
educated guesses based on his own knowledge. That his random behavior and
the perfect knowledge needed for quizzes needed to converge.
BTW, I can't attribute it solely to the near constant quizzing, but this
has been a great class. It is significantly better than my class last
spring (also with BlueJ and B/K text).
Best regards,
--
Jay Fenwick
Assoc. Professor, Computer Science, Appalachian State University
107-E CAP Science Bldg, P.O. Box 32133, Boone, NC 28608
jbf at cs.appstate.edu http://www.cs.appstate.edu/~jbf 828-262-2708
On Thu, 20 Apr 2006, Gordon Royle wrote:
> Yet quite a few of them just don't seem to grasp the concept that
> THEY are in control and that programming involves first thinking HOW
> to do something, and then DOING it using the tools available. Instead
> they seem to view it as some sort of randomized jigsaw puzzle where
> the goal is to shuffle around the pieces of code until it somehow
> magically works...
>
> How can I overcome this?
More information about the bluej-discuss
mailing list