Finalising survey questions for first tutorial delivery

[dhavide]

Feedback on the assessment questions:

Let's approach this the following way: the current assessment questions look like this:

• Preliminary questions: six questions;
• NASA EarthData: four questions; and
• Post-tutorial survey: two questions (a matrix-style question and an open-ended question)

Assuming this format, the goal is to choose questions that solicit the most useful information within that fixed budget of questions.

With that in mind, my feedback is this:

• The third section (the post-tutorial survey) looks good, i.e., those questions are fine.

• In the NASA EarthData section, these two questions ask nearly the same thing:

  • "Have you accessed NASA EarthData Cloud before this tutorial?"
  • "Have you used NASA EarthData products in your research or projects before this tutorial?"

  Could one of those questions be replaced by another that would solicit greater information? Perhaps the second one could be replaced by "Have you used any kind of GIS tool or products in your research or projects before this tutorial?". @JFormoso, please let me know that the reasoning is that I'm missing.

• For the first "Preliminary questions" part, it's challenging to choose questions. Assuming we want to stick to a fixed budget of six questions at most, again, which questions provide the instructors with the most useful information beforehand? On reflection, we want to assess the depth of their geospatial knowledge and the depth of their Python experience as much as possible (say, three questions each).

Designing good multiple choice questions is difficult; it's hard to choose meaningful distractors. I suggest we ask six short answer questions for the first section instead to find out what they really do know that could generate good distractors. @JFormoso, if you are in agreement with that suggestion, @kvenkman & I will come up with a list of six question to put into the Google form before Saturday.

A final complication to consider is that the SciPy audience will be somewhat stronger with respect to programming skill and probably geospatial expertise than the generic audience for this tutorial. So we'll have to decide how generic these questions should be, i.e., would we want to use the same questions for other audiences as well?