Clark Quinn's Learnings about Learning
(The Official Quinnovation blog)

5 October 2017

So I was, at least partly, wrong

Clark @ 8:08 AM

A number of years ago, I wrote that pre-testing learners was user abusive (with a caveat). My argument was sensible, qualified, but apparently wrong. Now that I’ve more of the story, it’s time to rectify my mistake. Of course, there are still remaining questions ;).

My claim was that while pre-testing might have some small benefit, forcing users to test on things they don’t know isn’t nice.  Moreover, I attributed that benefit to activating relevant material, and suggested that there were more humane ways to do it. However, if the pre-test could show that learners did know it, and so be able to skip it, it’d be worthwhile.

However, research has now shown more benefits to pre-testing. That is, causing learners to search for information they don’t have somehow makes the memory traces more susceptible to successful learning subsequently.  Without a full neurological explanation, it appears that the activation goes deeper than just associative awakening. It also appears to be for more than just memory, but actual performance.

This, then, argues that pre-testing is a good thing. Now, I haven’t been able to find a comparison where this pre-testing was compared to a compelling story or question that didn’t require an actual response. Still, I’m willing to believe that the actual requirement for search in a test is more powerful than mere related stories.

And this also makes the case stronger, in my mind, for problem-based learning. That is, if you’re faced with a problem you don’t know the answer to (and it’s a comprehensive question representing the overall learning goal), both the need to look for the answer and (ideally) a compelling story in which it’s important make a good case for the learning to be more effective.

Which doesn’t mean I don’t still feel it’s abusive, but it’s in a good cause.  And it still could be that the learner doesn’t actually have to take a ‘test’, but instead in some less formal way is asked to retrieve the answer.  And it might not.

Regardless, I feel obligated to change my opinion when data contravenes, even in part, a story I previously believed. And it doesn’t even hurt much ;).  Here’s to good design!


  1. Huh. Interesting.

    I can see how the connection makes sense (which of course has no bearing on whether it ends up being true. But it makes me feel better.) Seems like telling a story and asking a few questions without answers at the outset would likely have the same effect – motivate the learner to seek answers to questions – without the abusive framing of a test they’re bound to fail.

    Comment by Rob Moser — 5 October 2017 @ 9:43 AM

  2. You are always the learner and I thank you for updating. I have been in favor of pre-testing as a way to focus time efficiently but your share was added bonus. Competency-based design focuses on demonstrating in the context of the content to be taught so solving the problem targets the learner and the learning.

    Comment by William Ryan — 8 October 2017 @ 2:28 PM

RSS feed for comments on this post. TrackBack URI

Leave a comment

Powered by WordPress