Laura mentions a KLA she did as a student involving throwing a ball as a metaphor for throwing exceptions. (Might be neat to do this as rolling a ball down a track with people lined up as the stack trace along the track.. that way, the person throwing the exception doesn't necessarily know who's going to catch the exception. If no one catches it, it can fall off the end :) ... Maybe could use silly putty instead of a ball, accreting a stack trace in different colors of putty as it works its way up the stack?
Tammy says she thinks KLAs are valuable b/c they're novel (rarely used currently). Laura suggests that they might be valuable b/c they make abstract processes concrete. Monica points out that people retain more of what they do (rather than listening, reading, or just thinking about it).
Tammy describes the acrobats exercise.
(KLAs achieve "concreteness" not just from being physical activities but from exploiting analogies between common social knowledge (e.g., the distinctiveness/boundary between individuals or eye contact) and CS concepts (e.g., encapsulation of methods or pointers. Of course, this is also a danger if the limits of the analogy aren't explored. Otherwise, they can take away extraneous or inaccurate information from the analogy.)
Someone (Tammy?) describes the network routing exercise. She points out that a critical element of this, and other KLAs, is to have a discussion at the end to assess whether the students learned from the exercise, to encourage reflection to help them learn, and to explore the limits of the analogies in the exercise (as discussed above).
Later, David mentioned some specific reflection questions he'd seen used. Here they are from e-mail from David:
Here are the retrospective questions I mentioned last week at 590ed. They are from an editorial "Teambuilding at Work" by Johanna Rothman in STQE Vol 5, Issue 4, July/Aug 2003, p. 64:
- What stood out for you?
- How was this activity for you?
- What insights does this give you?
- What is one thing you will do differently?
Tammy describes a classwide, hand-simulated, distributed Monty Hall w/dice to decide door.
Steve describes the 10-person, front-of-class Monty Hall he did in CS1. (Students hide either a prize (cookie) or nothing by cupping hands and turning around. Another student plays the contestant. The larger number makes the logic of the puzzle clearer.)
Vibha describes a "find the lowest-priced CD" exercise (among other tasks) w/physical computer (post-it notes and such); different groups have different task so one group executes another's instructions w/out knowing the task. TODO: This should be added to the KLA site (Steve).
Vibha and Tammy mention Luke's human robot exercise (find cheese given a limited communication language). One can imagine doing it blindfolded for extra fun :)
Tammy describes a hand-simulated, distributed sorting timing exercise similar in spirit to the Monty Hall simulation mentioned above..
Janet describes an exponential backoff algorithm for sending which simulates the backoff algorithm with coin flipping. TODO: This should be added to the KLA site (Steve).
Steve describes an exercise in which a CS1 class just learning the concept of algorithm brainstorms different ways to count the number of people in the class, how fast they are, how accurate they are, how many resources they require, etc. Students came up with upper bound techniques (e.g., check the seating capacity of the room) on their own. Also, finished with a randomized algorithm which went awry in amusing ways: students all stand, think of a number 1-4, all who aren't thinking of 2 sit down; remainder think of a color red, green, or blue, all who didn't pick blue sit down; count the remainder and multiply by 12 (which was a vast underestimate perhaps b/c of the prevalence of choice of "3").
David describes a complex, extended activity at a workshop on team-work/management in industry (not necessarily CS). Very cool exercise! (Among other things, this illustrated the "mythical man-month" concept very concretely.) The product being produces is sorted decks of cards. For the MMM exercise, groups of two people sort two decks of cards and present the finished product (small companies/teams). Then, the groups of two are merged into groups of four w/four decks, etc. Scaling problems arise, understanding ensues :). Also, many other possibilities. David mentions a discussion of quality assurance based on first a question about how the groups know the decks are sorted and later the facilitator quietly swapping a card's location in the deck without being noticed (the point being that they hadn't clearly marked the product as finished).
TODO: Ken says he owes an exercise on garbage collection.