Didaktics of media: Evaluation of Instruction and On-line Exercise Design

This blog post is one part of the series “learning diary” for a course at my university.

Goals:

Based on the title and the provided pdf for that unit I assume we will learn how to evaluate and compare different examination methods for different kinds of learning goals. I think we will consider target groups, the goal of the examination and several other aspects of online examination.

In the end I hope I’ll be able to create my own online examination because we need to create multiple choice tests later on.

What have I learned:

On this lesson I only wrote a single sentence as comment in my notebook:

Grading on evaluation is always highly subjective

Reason for not writing more notes was the more or less complete slide my professor provided. So almost everything was covered in the slides.

So what have I learned and is covered on these slides? I learned the up and downsides of creating good online examinations. “Good”, in this case means several things. A good (online) exam needs to to take in consideration some organisational stuff (identity check of the exam taker, for example) and also needs questions which are precisely written and good to grade.

I learned of seven different question types for only exams:

  • Fill-in-the-blank
  • Essay
  • Ordering
  • Matching Items
  • Extended Matching Items
  • Multiple Choice (MC)
  • Multiple Answer

We had a quick overview over all of these types (especially from the point of view of someone who grades these tests) but went in more details with Multiple Choice Questions because we will write ten Multiple Choice Questions as an exam.

So, what learned I about MC?

  • An MC has one question and several alternative answers.
  • Only one of the answers is correct. The other answers are sometimes called distractors.
  • A MC with only to possible answers (“true” or “false”) is a special case.
  • A MC is easily scored and the marks can be evaluated by machines.
  • For all but the synthesis level of Bloom’s taxonomy, MC can be easily created.
  • The creation of a good MC can take more time than other kind of exams.
  • There is no way one can justify a choice (this is important if the answers aren’t as precise as needed).
  • Good MC Questions
    • shouldn’t give clues.
    • shouldn’t use jargon.
    • need to be as precise and simple as possible.
  • Good MC Answers
    • should be equally plausible. A set of answers to a question “What is the darkest color?” should only contain colors, for example…
    • should have the same length
    • should avoid “None of the above” and “All of the above”

Update 22.11.2015: Added “What have I learned”