shmdgljd@rchvmw3.vnet.ibm.com ("Jay Schmidgall Rochester, MN") (04/26/91)
Greetings. I don't know if this is the right group to ask so pointers and/or answers will be appreciated. Please reply to me direct at shmdgljd@rchvmw3.vnet.ibm.com Thanks. Also, I'm basically a novice with C++ so please excuse any obvious blunders. More thanks. Background: I've in an expert system class and we have to do a protocol analysis for our current assignment. To do this, I first need a protocol. A protocol is a trace, usually verbal, of how someone solves a problem, i.e., thinking aloud and having this recorded. The experimenter can then type up this trace and analyze it to hopefully find out the underlying method which is used to solve the problem. I would like to do this for task for someone identifying a C++ class based on some requirements. The requirements would either be a description of the desired class or an iterative process, e.g., providing some requirements, suggesting a class, accept / reject and add state additional requirements and repeat. Problems: 1) The C++ libraries that are available here are relatively straightforward, i.e., basically just the primitive classes such as stack, queue, iostream, etc. 2) Perhaps because of these simple classes, I am having problems coming up with a task to give to a subject. I can give her a description of a stack and she just goes, ``Well, that's a stack.''. Not much to analyze there. The iterative approach ends up to basically just a traverse of the class hierarchy. Requests: 1) Someone with experience with more complex classes than those I mentioned to give me some idea of what those classes are like. This would help me formulate a better task for the subject to solve. If you feel like it, propose a possible task description yourself. 2) Someone I can give a task to and have them record their thinking process. Due to time constraints, voice recording probably won't work. I was hoping someone with a slow day would be able to type up what they did to solve the task. I realize this is quite a stretch but hey, you never know until you ask :-). 3) Any hints on how to better go about this. Even more thanks. DDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDD Jay Schmidgall &standard_disclaimer. Internet:shmdgljd@rchvmw3.vnet.ibm.com