Five Years later: A Review of Kirschner, Sweller and Clark's Why Minimal Guidance during Instruction Does Not Work

After having a short discussion with Guy Wallace on his blog, I decided to do a review of Kirschner, Sweller and Clark's, Why Minimal Guidance during Instruction Does Not Work: An Analysis of the Failure of Constructivist, Discovery, Problem-Based, Experiential, and Inquiry-Based Teaching, in which they postulate that students who learn in classrooms with pure-discovery methods and minimal feedback often become lost and frustrated, and their confusion can lead to misconceptions.

The paper caused a bit of a stir in the learning and training community when it was published five years ago, especially among those who lean towards a more constructivist approach. However, while the author's critics did raise some good points, the paper is a good reminder that learning and training professionals often carry new ideals and technologies to the extreme. For example:

  • We had the visual movement from about 1900—1950, which brought us Dale's Cone of Experience. And of course someone had to add some bogus percentages to it to make it more “official.”
  • When VCR's arrived we made training tapes of everything… even if it did not make sense.
  • eLearning was supposed to kill the classroom.
  • Formal and informal learning were supposed to be at odds which each other, even though each hour of formal learning spills over to four-hours of informal learning.
  • All learning is social! Uhh… no. While the majority of learning may be social we often still learn things on our own.

Thus Kirschner, Sweller and Clark's paper is an important reminder for us to not carry Problem Based Learning (PBL) to its extreme. That is, while it has its strengths, learners often need a more direct approach in order to build a solid foundations before being presented with PBL.

With that being said, we do need to take a closer look at the paper. For those that are interested, there is a list of papers that discuss the Direct Instruction versus Constructivism Controversy (they are located at the bottom of the page).

The Title and Paper gives Little Respect to the Constructivism Approach

With the title blaring, “Why Minimal Guidance during Instruction Does Not Work” rather than, “Why Minimal Guidance during Instruction Does Not Work for Novice Learners,” the authors almost seem to ignore that PBL is a necessity in order to promote deeper levels of understanding. They do pay some respect to constructivism, such as:

Higher aptitude students who chose highly structured approaches tended to like them but achieve at a lower level than with less structured versions

Certain aspects of the PBL model should be tailored to the developmental level of the learners… there may be a place for direct instruction on a just-in-time basis. In other words, as students are grappling with a problem and confronted with the need for particular kinds of knowledge, a lecture at the right time may be beneficial.

However, they end up admonishing constructivist:

According to Kyle (1980), scientific inquiry is a systematic and investigative performance ability incorporating unrestrained thinking capabilities after a person has acquired a broad, critical knowledge of the particular subject matter through formal teaching processes. It may not be equated with investigative methods of science teaching, self-instructional teaching techniques and/or open-ended teaching techniques. Educators who confuse the two are guilty of the improper use of inquiry as a paradigm on which to base an instructional strategy.

But it seems, at least to me, they may be doing the same, but only at the opposite end of the continuum. For example, they seem to treat their theories as laws, yet…

Cognitive Load Theory Coming Under Withering Attacks

The paper relies heavily on Cognitive Load Theory, yet we have to realize that it is still a theory rather than a law. Will Thalheimer lists several papers on his site that raises several concerns about Cognitive Load Theory. For example, even though we know that working memory can only hold about seven chucks (which actually may only be four, give or take one), using the old KiSS (Keep it Simple Stupid) principle can be just as effect because trying to count the number of chunks can be quite difficult, if almost impossible. For example, how many chunks are in Rene Descartes statement, “I think, therefore I am?”

Thus, both the authors and the constructivism movement are guilty of jumping on theories before they are fully understood. But why do we do this? Joel Michael writes in Advances in Physiology Education:

…it is important to recognize that educational research is difficult to do; this has been cogently highlighted by Berliner (8) in "Educational research: the hardest science of them all." Berliner points out that unlike a physics experiment, in which it is possible to readily distinguish between the independent and dependent variables, and also possible to isolate and control all of the independent variables, in educational experiments all of this is problematic. Researchers may not agree on which variable is the dependent variable of greatest interest or importance. There may be disagreements about which independent variable(s) are to be manipulated. There may be disagreements about how to measure any of the relevant variables. And, finally, it may be extremely difficult, or even impossible, to isolate and manipulate all the variables suspected of being involved in the phenomena being studied.”

Rather than waiting for eons to pass before all the research is available, we (the learning, training, and educational community) often jump into a new theory because will simply do not want to wait until we are dead and buried before we can fix and/or improve our methodology. With that in mind…

Evidence for Constructivism

Joel Michael continues his discussion for promoting active learning with these two studies:

1. Support for discovery learning is provided by a study in which students engaged in a course that incorporated some discovery learning exercises were tested, and their performance on questions related to topics learned through discovery learning was compared with their performance on questions related to topics learned in lecture (Wilke, Straits, 2001). The authors concluded that performance was better on those topics learned through discovery learning.

2. Burrowes compared learning outcomes in two sections of the same course taught by the same teacher. One section was taught in the traditional teacher-centered manner (control group of 100 students), whereas the other section was taught in a manner that was based on constructivist ideas (experimental group of 104 students). The results of this experiment were striking: the mean exam scores of the experimental group were significantly higher than those of the control group, and students in the experimental group did better on questions that specifically tested their ability to “think like a scientist.” Reference: Burrowes PA. Lord's constructivist model put to a test. Am Biol Teacher 65: 491–502, 2003.

While you can find plenty of other research findings on constructivist methods, the ideal that you can teach learners to “think like scientists” is fascinating because problem solving skills are extremely hard to train. That is, conduct a problem solving course in an organizational setting and you will more than likely get little or no results. It's almost as if the process must be embodied within the discipline.

Embodied Cognition

On the Brain Science Podcast, Ginger Campbell discusses Embodied Cognition with Lawrence Shapiro (both podcast and transcript can be found in the link). They note that in cognitive science, the brain is normally studied while isolated from the world and from the body. While in contrast, embodied cognition imagines not that the brain can be isolated from the body and the environment, but thinks of the body as in some sense shaping, or constraining, or involved in the very processing of the kinds of information that an organism needs to interact successfully with the world.

In the podcast, Dr. Shapiro talks about a fascinating work on the use of gesture. He notes that boys perform better in certain spatial reasoning tasks than girls. When psychologists studied this they've noticed something kind of interesting—boys rely on gestures a lot more than girls do when solving spatial reasoning tasks. Boys use gestures to work out the problem, at the same time they're talking through the problem. And often, they don't synchronize with the verbalizing. It's as if they have two different systems working at the same time—one a gesture system, and one a verbalization system. Gestures seem to be a part of the process of figuring out these spatial reasoning tasks.

They also discuss the study of kittens moving around their environments by pulling carriages with other kittens in them. And the ones in the carriages presumably see everything that the kittens pulling them see, but because they don't actually employ their motor systems to move them around the environment, their perceptual systems don't develop properly. The idea there seems to be that part of what's necessary for perception is actual exploration of the environment; not just being a passive recipient, like these kittens conveyed in the carriages are.

Thus rather than focusing on pieces of conceptually unrelated pieces of information, such as practice first and then learn problem solving second, perhaps we should be focusing our learning processes on entire ideas and concepts whenever possible.

I would interested in your thoughts on the subject by leaving a comment, creating a blog post, or through Twitter (I'm @iOPT).


Anonymous said...

I want to come back and read this more thoroughly (it's late for me, and I'm out of the office the next few days). One question / notion floating around my mind is:

Given that a lot of this research seems to involve learning in formal education settings (grade schoolers, college students), what are the implications (if any) for workplace learning -- where the instructional control (so to speak) is less and where the participants are in theory more in charge of their own actions?

I don't have any agenda in mind, just wondering out loud.

Donald Clark said...

Hi Dave,
I believe we have the control, but we don't often use it. Kids go to the school because of their parents and family. In homes where there is little or no parental support, kids often do worse than their peers.

Teens go to college because they see a good reason behind it.

Peers also can support or hinder the learning process.

One of the equivalents in the workplace are the managers--employees most often do what their managers stress. Thus the managers need to discuss with their employees what they expect before they go to training and follow-up when they come back. Of course the employee should also discuss what he or she sees as important in the learning process. So its really more of a reciprocal contract.

By including some of the employees in the design process you can build peer support (we like to be part of the change process, not merely a reaction to it).

This of course means that the training has to have a real purpose in order to build this support group.

JLF said...

Thought provoking as usual. I believe motivation to learn and interest in the subject are paramount. What do you do in a work-place environments where you are dealing with novices to the subject matter, and perhaps under-developed professional skills, such as punctuality, and self-direction? I'm frequently distressed when we expect our hourly employees to be "professional," but the work environment & management treat them like errant teenagers.

Donald Clark said...

This sounds more like an organization culture issue than a training issue so there is no short answer. Let’s start with the punctuality issue by seeing if the following is helpful.

What normally happens in a classroom when you have late arriving learners? We wait until the late learners arrive. Thus we punish the good behavior by wasting the learners’ time who arrived on time and punish ourselves by making us rush though our lesson plans. In addition, we reward the bad behavior by showing we will wait for them. We are using positive reinforcement, but in exactly the opposite direction than it should be! So what we should be doing is rewarding the good behavior by greeting them at the door and then starting on time to show we appreciate their timeliness. And we punish the bad behavior by making them catch up on their own time, but as professionals we will help them.

This may not solve the organization culture issue, but at least it will bring some peace and proper reward to the classroom. In order to extend it to the workplace we can start with some training, but not to the employees, but rather to the managers and supervisors by teaching them the above technique of positive reinforcement. In addition, we can train them on Confrontation Counseling for the employees who still don’t get it - See http://nwlink.com/~donclark/leader/confront.html

Of course we need to bring someone on board who can help us – the manager in charge of the other managers and supervisors (such as the plant manager), but first we need some evidence that the training will be worth it. So pick a few random work locations and see how many employees are showing up late and the average time they are late by. Calculate their average wage and then determine the lost wages by pay period and what it takes out of the annual budget (these two time periods are near and dear to most managers’ hearts). Also note that this only accounts for the lost wages. The other thing that is lost is the productivity gains that could have been made if the employees were doing what they were supposed to do!

Now determine if the punctuality issue is worth the training effort or if it is just a small issue. If it is big, then set up a meeting with the manager and get her buy in for you to not only perform the training, but also her help in discussing with her managers what they should expect to do differently once receiving the training and following-up when they come back from training. And of course the managers receiving the training should also discuss what he or she sees as important in the learning process - a reciprocal contract.

You of course should also be in the follow-up to see if there are any obstacles preventing the managers from putting their new skills to use (such as they are in meetings during start time, which prevents them from performing the positive reinforcement). In addition, you might learn that they need more training or that you could have done something different (this is a learning opportunity for you too!)

@siibo said...

Several semi-random-to-you-but-connected-in-my-head points:

I'm not a fan of 'Multiple Intelligences'™ (with capital letters) as a theory but the idea that we may all have different ways of achieving success seems almost trivially true. For something to be 'universally true' there doesn't necessarily need to be a solitary single way of doing things, but there should be something bell-curve-ish with a single median/mode point. I don't think intelligence fits this pattern.

Likewise, Learning Styles™ is also bogus. But, again, the fact that different people favour different modalities is trivially true (even if the impact of how we work with/against these preferences is potentially a minefield).

And then we have scaffolding. After the age of about a day, all there's simply no way of controlling for or predicting the scaffolding we draw on when we learn - whether this is the stuff we've internalised or the stuff that's distributed/embodied in our environment.

I could go on - but the sheer number of potential salient variables is astonishing. And splendid.

I think we've been over this before, but I'm not even sure this is something that will ever be solved theoretically. Small bets, 'agile' planning and data-driven (not necessarily numbers - calm down epistemic hippies!) feedback loops might (might!) always trump anything driven purely by a learning theory.

But we need the theories because we need to scale 'learning' in organisations and in educational environments.

I guess what I'm trying to say is that many learning professionals seem to see each new paradigm that comes along as a battle to be fought as if they can't all co-exist. When they can.

If we take, say, Constructivism vs Direct Instruction, it seems perfectly plausible to me that any independent learner might reach a certain point and select chalk-and-talk as their preferred learning strategy. Constructivism contains the 'previous' paradigms as much as it competes with them.

To be honest, I think all that I've said is fairly innocuous. And where's the fun in that?

So, to finish, I'll make a brash, ill-thought-out statement:

The people who cleave to one or the other theories struggle to deal with ambiguity. Not all domains are characterised by ambiguity and there are plenty of trainers who do invaluable work teaching people how to navigate and perform in such environments. But, in general, if you can't tolerate ambiguity and find yourself passionately on one side or the other of any Constructivist vs Direct Instruction (for example) debate, you should get a job in another field.

And relax.

Guy W. Wallace said...

Excellent discussion! I too need more time to read these comments and responses over again. But thank you for your Post!