So, as an instructional designer, part of my job is to make things clear and easy to understand, right?
Well, it turns out that’s not necessarily the best option.
Cathy Moore just put up a blog post that has her checklist for evaluating your own e-Learning design. You rate where your learning falls on a continuum. In particular, I noticed this item:
This isn’t a new idea, but it’s particular powerful one — use consequences instead of disembodied-voice-of-the-eLearning-gods-type-feedback.
Sure, the “correct/incorrect” feedback may be easier to understand or have no possibility of misunderstanding, but it’s a disservice to your learners (and not just because it’s boring).
It’s been particularly resonating with me because of something that was said in this podcast on Show, Don’t Tell for fiction writers (mp3 here) from Storywonk:
“The difference is that in Telling there’s absolutely no role for the viewer or the reader to put anything together…In Showing, the viewer has a chance to put two things together…it’s giving them the opportunity to put stuff together themselves and to actually be active in the story…”
“It’s so much more engaging as an audience member… if I am left to put stuff together myself and not have it all assembled for me and handed in front of me that this is the way it is.”
“You need to give your…readers stuff to do. Give them a way to be an active participant, and by allowing them to draw conclusions based on little clues that you leave, you engage them in the story and they become part of [it]…”
– Lani Diane Rich (aka Lucy March)
(emphasis and any transcription errors are my mine)
I thought that was a really interesting take on the issue. From a learning point of view, reading a text is considered to be one of the more passive ways to learn, but your text can be really passive (let me just hand everything to you ) or it can be made more active through showing rather than telling.
I think this matches up really well with the point that Cathy is making. It’s one thing to say “It’s really important for health care practitioners to wash their hands” and entirely another to uncover the fact that a horrible staph infection is threatening vulnerable patients in the hospital.
A little friction is necessary for learning. Making something very easy to understand is actually doing your learners a disservice. I just saw this fascinating critique of the Kahn Academy videos (found via the Action-Reaction Blog). In it, Derek Muller explains that the “easier to understand” version of a science video had worse outcomes:
[youtube=http://www.youtube.com/watch?v=eVtCO84MDj8]
Learners who heard “clear and easy to understand” explanations did worse than students who were confused by discussions of misconceptions. In fact, learners from the “clear and easy to understand” camp frequently thought they’d understood when they hadn’t (watch the video – it’s really really good).
This goes along with the incredibly interesting study that came out of a few months ago that looked at this question (I paraphrase):
What’s the best way to study for a test?
a) Read the text
b) Read the text in consecutive sections
c) Create a concept map of the material (described in the NY Times as “arrang[ing] information from the passage into a kind of diagram, writing details and ideas in hand-drawn bubbles and linking the bubbles in an organized way”).
d) Retrieval practice (a free-form essay test followed by re-reading, and a second test).
Dave Ferguson has a good write up of this study with a link to the actual paper, but the test-taking condition beat the other hands down. They were even better at concept-mapping the material a week later than the students who had actually been part of the concept-mapping group. The researchers speculate that it’s partially because the learners were forced to confront their own knowledge gaps and reconcile them rather than just recognizing the material and assuming they knew it.
Another interesting perspective on this is from this study: Making sense of discourse: An fMRI study of causal inferencing across sentences
Subjects were shown sentence pairs. Some of the sentence pairs went together very easily (x obviously causes y), some required some interpretation to see the connection, and some were pretty unrelated. For example:
Main sentence: “The next day his body was covered in bruises.”
That sentence was preceded by one of these statements:
-
- “Joey’s brother punched him again and again.” (highly causally related – x obviously caused y)
- “Joey’s brother became furiously angry with him.” (intermediately causally related – you’ve got to read between the lines a little)
- “Joey went to a neighbor’s house to play.” (pretty much unrelated)
The subjects spent the most time on the middle sentences — they were related but forced the subjects to connect some dots to see the connection. The study saw a greater degree of brain activation in many areas for those sentences, and they were better remembered later.
So, in the end, there appears to be something really beneficial about wrestling with stuff a bit and drawing their own inferences — you need to have a certain amount of learning friction.
I’m not arguing you should make things deliberately obtuse (there’s a difference between challenging and confusing), but if learner can connect the dots too easily, they don’t retain the learning as well (or as Derek Muller points out — they may think they DO know when they really don’t).
And if you can create opportunities for the learners to confront their own assumptions, and give them access to their own gaps, the overall results will be much better.
Whaddya all think? And any ideas for good ways to add a little friction?
—————————————————————-
As an aside, this has interesting implications for Level 1 Evaluation (Level 1 = What was the learner reaction? Usually interpreted as “Did your learners like it?). It suggests that a positive learner reaction (“It was clear and easy to understand!”) can actually be a counterproductive measure in certain circumstances. Hmm.
References:
– Karpicke JD, Blunt JR (2011): Retrieval Practice Produces More Learning than Elaborative Studying with Concept Mapping, Science 11 February 2011: 331 (6018), 772-775.
– Kuperberg GR, Lakshmanan BM, Caplan DN, Holcomb PJ (2006): Making sense of discourse: An fMRI study of causal inferencing across sentences. Neuroimage 33:343–361.
– Muller D: PhD Thesis, Designing Effective Multimedia for Physics Education, http://www.physics.usyd.edu.au/pdfs/research/super/PhD%28Muller%29.pdf
Great observation. It would be interesting to try this with adults and different topics. Maybe compare adults who are required to complete the training to keep their jobs vs. adults who pay for training to update their skills or advance in their present careers.
Yeah — I think you’re right, Jill — obviously intrinsic vs extrinsic motivation has a role to play here. If nothing else, it relates to their willingness to make the effort presumably.
Good post, and nice summary of some interesting studies. Robert Bjork has advocated the concept of “desirable difficulty” for some time now, and his studies consistently show that learners overestimate how much they learn when material is presented in an easier way. Forcing them to do just a bit of mental juggling results in astonishing increases in actual learning.
Thanks! And thank you for the pointer to the “desirable difficulty” material — I want to do a longer article on this topic, and I will definitely check out that reference.
Your comment also helps clarify the way I’ve been thinking of it — I don’t think I made it that clear in the post (I’m not sure it was clear to me as I was writing it!), but I think there are two big issues at play 1) that mental effort helps consolidate memory and 2) that learners tend to be overconfident of their own understanding unless there’s some way to help them see their gaps.
Thanks for the comment – very helpful.
Wonderful! I think it was Dan Meyer that called it being “strategically less helpful”. Hand-holding is done with the best of intentions, but still robs learners of the opportunity for deeper understanding, retention, recall, transfer, etc.
I want the pilot of my next commercial flight to have gone through some horrific conditions in a simulator, and I hope the pilot was given the chance to crash rather than simply having it all neatly explained.
All that said, though, with difficult topics (like computer programming), we see far too many examples of learning materials that are NOT clear and easy to understand, but not from strategically crafting the scenario where the learners will connect the dots… they are just confusing and hard because they are created by a SME who forgot what it was like to not know, etc.
So, I still put clear and understandable as the prereq, and THEN start figuring out what, when, and where to hold back, and building scenarios that best support the dot conmecting *without just pissing everyone off.* I think it was Mamet that said, “if one character is talking to another about a third, the scene is s***” in reference to the awfulness of exposition (tell vs. show).
It takes a bit of bravery to use this approach since you DO risk frustrating people, so it is not for the less-skilled knowledge designers (including book authors), but it is like a superpower when used appropriately 🙂
I love the Dan Meyer “be less helpful” stuff. I think he talks about it in terms of learning the skills necessary to approach messy, ill-structured problems, but it definitely applies here as well.
I completely understand about so much material being confusing and hard just cuz it’s bad, not by design! It does make it a little touchy to criticize things that are too easy to understand, because there’s so much stuff out there that doesn’t meet even that standard.
It does require bravery to do — it’s such a counter-intuitive idea. I was watching the MIT Gambit keynote yesterday http://gambit.mit.edu/ and they were talking about a game they designed for the Smithsonian. It sounded like an ARG, and was very much based on emergent game play — they didn’t lay it, and provide instructions and have a highly facilitated experience. Instead they laid clues that could look like system glitches, and left messages with puzzles. The kids managed to figure it out through knowledge sharing on the forums, which sounds great. Still, I had a little bit of a “but, but, but…” reaction (even though I know better) about the kids who might not get it, or might be left out if it wasn’t made clear. D’oh!
I’m one of those people who gets anxious if I give directions to someone on the street (What if it was confusing? Oh no!), so that leap of faith to trust learners is something that I have to consciously do.
(on an unrelated note, I have a couple of things I need to ask you about, Kathy. Will send in email)
Great stuff! Reminds of the articles back in January when the research about disfluency came out. Like this one: http://www.wired.com/wiredscience/2011/01/the-benefit-of-ugly-fonts/
Also reminds of a user test we did years ago where the users got upset about getting quiz questions wrong. They had wanted the obvious “what did you see three pages ago?” in order to keep up appearances of knowing everything (they weren’t SME’s). In their defence, I think we probably had “that was wrong because…” feedback then…
Yeah, the disfluency is interesting stuff (thanks for the reminder!) — it’s similarly counter-intuitive, and I suspect it operates in a similar fashion (the extra effort is what causes increased retention).
Interestingly, when we look at learning to change behavior, the disfluency stuff has some interesting ramifications. I can give you the instructions for a new exercise in a harder to read font, and you might remember those instructions better, but based on other studies, you’re also likely to perceive the task as been more difficult to do, which may have a negative effect on the likelihood of actually doing the exercise.
Humans are wacky creatures…
Uhm, someone has “Good Job! You correctly identified Option A as the correct answer. That is correct!”, in a checklist advocating consequences over correct/incorrect? I’ve got some cognitive dissonance going on now. I’ll have to check back when i’ve cleared it.
On a related topic, how do you teach students to keep thinking past the minimum requirement? What hints can you give them for what to think about? My list for math might include looking for a way to check the answer. If you’re solving for x, does your solution actually work? And, is there another way to solve the problem? There must be lists for other subjects. I swear that we’re breeding ADHD. (Oh, look, bright, shiney!) Or is it Star Wars training (“Live in the moment”)?
Rumsfeld said “You don’t know what you don’t know”. And he was wrong. Apparently not all tautologies are true. I didn’t know that. Thanks, Donald. But it is nice to get a hint now and then. The student has to do the work, but without proper hints, the student is left in the stone ages to build civilization by herself. There’s got to be some middle ground between “Here’s the formula for the area of a square” and “Derive a formula that gives the area of a square”.
Stephen, the ‘Good Job’ graphic was my visual example of “disembodied-voice-of-the-eLearning-gods-type-feedback” – sorry if that wasn’t clear (and it’s not from Cathy’s stuff at all – I’m solely responsible).
Regarding your second comment, Kathy mention Dan Meyer in her comment, and I’d refer you to his blog (http://blog.mrmeyer.com/) — he’s a master at walking that line of giving just enough information, but not too much. This video is a really good explanation of his approach: http://www.youtube.com/watch?v=jRMVjHjYB6w
[youtube=http://www.youtube.com/watch?v=jRMVjHjYB6w]
Julie, though not a ux expert at all myself, I love your blog. And this post is terrific.Am a comms psychologist and one heuristic I’m always looking for in messages is vividness. In the Joey’s brother comparison, the most impactful sentence has the most vivid language ‘furiously’. This is well researched in social psychology lit, am sure.
Please continue to be so vividly interesting and look forward to your book
Thank you! So glad you are enjoying the blog.
I think the vividness comment is interesting — I’ve been interested in the importance of urgency and immediacy in gaining and maintaining attention, and I think vividness is a key part of that, although I hadn’t been using that term (very helpful).
The Joey’s brother example was the one given in the paper, but I’d actually like to see the other sentence pairs they used. Apparently the study has been done previously without the fMRI component, so maybe those papers have other examples. It would be interesting to see if that vividness distinction was just this example, or a consistent thread.
This is perfect. I have a project at this very moment that will benefit from the insights shared here. Thank you
This is an excellent description of the gaps that I believe we all face in our own learning. As a new I.D. grad student and in the early stages of my program, I am amazed at the level of understanding of learning theory that is needed to produce learning scenarios with effective outcomes.
I especially appreciate the comment that learners “were forced to confront their own knowledge gaps and reconcile them rather than just recognizing the material and assuming they knew it”. As a former high school teacher and now college faculty member, I feel like I employ these strategies without really understanding them. I do use the concept of “modeling” often in my classroom (and will as an I.D.) to help students make the kinds of connections you discuss, while asking them to employ their own scenarios to solve the problem. I find this inadvertently helps students confront their own knowledge gaps when they have to construct solutions based on concepts held together by their own beliefs and ideas.
After reading this post, I realize that while modeling is a form of friction, I would do well to harness this powerful tool and create more friction to help my student’s learning, especially when I will need to work without the use of learner constructed scenarios.
Thanks for your comment!
I think you are right — we do introduce “friction” in a number of ways in instructional design. I’m interested (for my own practice) in seeing how that changes my design decisions when I’m explicitly trying to create the situations where learners are “forced to confront their knowledge gaps”. For example, if that’s the instructional design goal of a problem-based learning activitiy, how does that impact the design? Interesting stuff to think about!
I enjoyed reading your blog post. I have taught special education for 20 years and know that clearer and easier is seldom as effective as letting the students problem solve to find the answers. Allowing the students to figure out for themselves how to complete tasks to come up with a reasonable answer is an easy way to differentiate lessons. My students soon learn they need to ask questions if they do not understand and to question if they disagree. As an adult learner I like to have control of my learning and I remember more when I am actively involved instead of sitting and listening to a lecture. Dr. Jeanne Ormrod’s visual clip “Information Processing and the Brain” explains that storage of information in long-term memory by relating the information to things you know and to situations in which you might use it. When told an answer is correct or incorrect without being given the chance to really understand, process and store it for later retrieval the information may be lost (Laureate Education, Inc., no date given).
In your blog post you stated, “A little friction is necessary for learning”. With this type of friction the student has to use problem solving and other higher level thinking skills to find a solution to a given task. There has been research on the effects of different learning situations many of which are similar to your clear and easy discussion.
Gregory Francom presented the paper “Involve Me and I Learn: Providing Substantial Learning Choices in Higher Education” at the Annual Meeting of the American Educational Research Association. In this paper Mr. Francom presents a report on how to increase motivation and learning in higher education. The theory is that if the student chooses the tools to complete an assignment and chooses the process to complete the chosen assignment he or she will learn and remember a great deal more. By engaging the learner and making him or her more responsible for their learning they are being actively involved and will remember more of what is learned. The research was conducted by giving a choice of what software, to use and what their project would look like when completed. Many of the students did not know anything about the software; some needed more teacher direction and visual clues while some thought they knew a lot more than they ended up knowing. As friction increased, the need for more information on how to use software or to proceed, the more the student wanted to learn and he or she spent more time and energy in their researching and the more knowledgeable they became on their topic (Francom, 2011).
A little friction is necessary for learning.
References:
-Francom, G.M. (2011). Include me and I learn: providing substantial learning choices in higher education. Paper at the Annual Meeting of The American Educational Research Association. Retrieved from Walden University Library, ERIC database.
-Laureate Education, Inc. (Producer). (No Date). Information Processing and Problem Solving. Baltimore, MD: Ormrod, Jeanne.
Carol — thanks for the references. I’m probably going to expand this post into a full length article at some point, so I’ll definitely check out these resources!
I really enjoyed this post on Why “Clear and Easy to Understand” can be bad. When I reflect on my own educational experiences, in every instance that I had to really
‘work’ or apply myself to grasp a concept, the more engrained it became in my memory. I believe that those learning concepts that we struggle with are more memorable. Whether the struggle forced me to better encode the information or it was a case of increased rehearsal and practice, I cannot say definitively. As an educator I reference Benjamin Bloom’s taxonomy as a model when creating lesson plans. This cognitive learning process is the model used to move my students from the lower-order thinking levels such as remembering to higher-order thinking levels like evaluating and creating. I find that facilitated discussions, activities which allow students to create and/or apply their knowledge helps to better encode the information for them. My goal is to use tasks that require metacognition. Clear and easy is used temporarily with my remedial students to build basic knowledge and then through exploration, have them continue to ask themselves “Why?”
Thanks for the thoughtful comment, Carol. I think your point about using clear and easy temporarily is entirely valid. I would never advocate for deliberately obscuring information, but I think that sometime we view our job as done when we get to clear and easy, and it’s really just the beginning (which agrees with what you are describing, I think). Thanks!
Hi there. I found this post via the pingback at Franck Noschese’s old Khan Academy commentary. I’m glad to see this theme — the value of “friction,” as you’ve call it — continue to spread. The problem I face as a math teacher at the college level is that confused students write negative teaching evaluations (even when the confusion was deliberate and temporary), and too often, college instructors are judged almost entirely by their teaching evaluations.
Your post is so very true in all age group of learners. I have taught world history in a high school in the metro Phoenix area for five years. Many students hated world history and I tried extremely hard to make it more interesting while making the students think more for themselves. With sophomores, sometimes in the beginning I had to do more handholding than I liked and then ween them off gradually as I taught them how to think for themselves. For instance, if I was going to have students label a map I would model on the board how I expected the students to label their maps. The next time we labeled maps, I would do a few examples and then have them work on their own and ask a neighbor if they were confused. The next time we labeled a map, I would have students work completely alone for the majority of the time and then allow then a few minutes at the end to check with a neighbor or myself for anything they were missing. Although the task of labeling a map may seem easy to adults and some of my students, around half of my students had really never had to label a map clearly in order to use it later. Once students were required to label the map by themselves they were able to recall the information they labeled more readily rather than just copying off the board.
When I think about my own learning experiences they are similar to several of your readers – the more they struggled and learned something the higher the likelihood of them remembering the topic. As an adult I am no different. When I struggled with my first master’s class because I did not fully understand APA citation styles, I had to work harder to triple check and fix errors in order to receive a better grade. Now in my second masters class, when I am writing and citing sources it is second nature to me and the citations are not slowing me down.
There are disadvantages of hard and vague explanation too. Such as talking about quantum mechanics to 4-years-old children, people which are not fluent in english, people who don’t know the basic of physics. KISS principle is useful in certain situation.
Absolutely agree, and as several people pointed out to me when I first posted this, ‘hard and vague explanation’ are a much more common problem 🙂