Teaching Tip: Testing..testing..testing

I’ve never been a huge fan of tests, preferring more “real world” tasks like projects and papers.  My attitude probably stems from too many memories of cramming for tests and then forgetting most of what I “learned”.  But research in learning and cognition suggests that tests can be very effective learning tools (not just assessment methods) if they are used well, so I’m working on readjusting my beliefs.

Regardless of how we feel about them, tests are a fact of academic life for most of us.  We all know that students pay more attention to material they think will be on the test (and as we all know, they quickly learn not to attend to material they think will not be on the test).  A truism in education at all levels is that assessment drives instruction; what you test is what students are most likely to learn.  So, our tests need to be as good as we can possibly make them.    I was pleased to come across a great summary of research-based learning principles that had many suggestions about what makes good tests.  (http://psyc.memphis.edu/learning)

Test Often
I still remember a comment Linda Elder of the Foundation for Critical Thinking made in a seminar I attended many years ago.  It went something like this:  We all know that students don’t study until the night before the test.  Instead of bemoaning this, use your knowledge.  Give them a test every class.  That way they’ll study at least twice a week.   While I’ve never gone quite this far, frequent testing has several benefits:

  • It keeps students more engaged with the content of the class and it slows down forgetting.
  • Frequent testing is particularly effective if you give students immediate feedback about their results and if you use the class results to guide your course planning.
  • How frequent is frequent?  Anywhere from every class to every two – three weeks.
  • I have not seen evidence that “pop” quizzes are any better or worse than announced ones.  I have not been a fan of pop quizzes since I was a student and don’t give them, but there is no pedagogical reason not to if that is your preference.

Start early

  • Spacing tests over the entire semester (rather than concentrating on mid-term and beyond) produces better retention than fewer tests, or tests that are closer together in time. Testing right after you present material usually results in higher scores, but long-term memory has not yet been engaged so retention is poorer.  We often think our students understand better than they do based on those first tests, only to be disappointed on the final or in the next course when the learning has disappeared.

Use constructed response items as much as possible

  • If you use multiple choice items, combine them with constructed response items (short answer or essay) as often as possible.  When students have to generate their own responses instead of recognizing the correct response from a list, decide true/false, etc. their learning is enhanced.  You can also include free response tasks as part of your teaching on a regular basis.

Avoid negative suggestion effects

  • A downside of testing is that students may remember the wrong answer they wrote or selected instead of the correct response.  To combat this, students need feedback as soon as possible on their performance so they can correct their memories.  You could do this by asking students to rewrite answers correctly or by reviewing items in class or individually.  One effective review strategy I have used with multiple choice tests is to have students immediately re-take the test in small groups with the group coming to consensus about the correct answer.  Then the class reviews any items that the groups got wrong (usually this comes down to just a few items).   This makes the test into a powerful learning experience, although it does take up more class time.  If you are requiring students to do more content acquisition on their own, outside of class, you will have time for this type of approach.  

Give a cumulative final (or at least threaten to)

  • When students expect a final exam that covers material from the entire course, as well as perhaps some more recent material, they will work to keep that material in their memory since they expect they will need it later.

Spreading testing out over the semester may have some benefits for you as well.  You can spread your grading out a bit more, you may be able to give more constructed response items if they are spaced through the semester, and you may see some happy improvements in student performance overall.

Oh and by the way, you don’t have to call them “tests”, and you don’t have to grade every one.  You can use clickers to pose questions and process them during class, or you can schedule “in-class directed writing” assignments every couple of weeks.   The important points are 1) frequent assessment of student progress; 2) make the students construct their responses as much as you can; 3) correct errors very promptly and 4) don’t let them learn something just for one exam and then be able to forget it.

Share your test giving strategies or concerns with us!


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s

%d bloggers like this: