1
$\begingroup$

I'd really like to ask a question about "best practices" in evaluating the performance of students taking computational science classes. Is there a way to do that within the current scope of the board? This would seem to be the best place to do it.

EDIT: The question is now posted.

$\endgroup$
4
  • 1
    $\begingroup$ How different is it from the question "what core skills should every computational scientist have"? I can't see a question about computational science students being very different from this. $\endgroup$
    – Paul Mod
    Mar 6 '12 at 4:33
  • 1
    $\begingroup$ My question is quite different, actually: "How can we best measure if students have acquired the core skills we are trying to teach?" $\endgroup$
    – aeismail
    Mar 6 '12 at 9:56
  • $\begingroup$ @aeismail: Isn't that a general educational testing question? Unless you can demonstrate how scicomp education testing differs from the problem at large. I don't think its topical. $\endgroup$
    – Deathbreath
    Mar 7 '12 at 13:25
  • $\begingroup$ Computational science requires resources that fall well outside the domain of what can be tested in a traditional pencil-and-paper exam. The methods for evaluation therefore are probably quite different, and also specialized relative to other disciplines. I discuss this in the original question. (You can see the question at the link above.) $\endgroup$
    – aeismail
    Mar 7 '12 at 13:49
1
$\begingroup$

My thoughts on such questions are as follows:

  • I think questions about best practices (in general) are great, as long as they don't ask for a list of best practices (canonical sources could be okay, if it's clear to experts that there's a small number of canonical sources).
  • I strongly oppose the creation of any "best-practices" tag because it is a "meta tag" (not to be confused with a tag on Meta, or a "Meta tag"). Jeff Atwood wrote a blog post on why meta tags are bad, and I agree with him.
  • Questions on teaching students, evaluating students, and self-learning have gone either way here. Good questions on these topics include How should I study creating and programming HPC systems? and Is algorithmic analysis by flop-counting obsolete?. Bad questions on these topics have been closed (and are likely to be deleted if they haven't been already). As long as you avoid bad subjectivity (see "Good subjective, bad subjective"), questions that don't have wrong answers, and being too localized, you're probably okay.

Of course, that said, you'd have to ask the question in order for me to have some idea of whether or not it's a good fit for the site, but having asked about it on Meta beforehand means that people who pay attention to Meta are much more likely to hear you out. (And you can bitch in this post if you think your question got closed unjustly!)

$\endgroup$
2
  • $\begingroup$ I've given it a try. Let me know what you think. $\endgroup$
    – aeismail
    Mar 6 '12 at 14:59
  • $\begingroup$ Looks good to me. $\endgroup$
    – Geoff Oxberry
    Mar 7 '12 at 2:11

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .