Monday, March 05, 2018

Why is online learning ‘all fur coat and no knickers’? Media rich is not mind rich

Online learning has gone down the ‘all fur coat and no knickers’ route. It’s more presentation than pedagogy, more look and feel than learning. Rather than focus on what makes learning a success in terms of understanding, retention and recall, it allows the learner to skate across the surface of a thin layer of crisply designed but thin ice. It often creates the illusion of learning by illustrative graphics/animation that, as Mayer often showed, actually inhibit rather than help retention. That old adage, which is as good a summary of learning theory as any, that ‘less is more’, has been abandoned by a glut of over-engineered graphics, animation and effects. We design for forgetting.

Rather than taking our lead from the most successful online services the world has ever seen, such as Google, which has the simplest, most successful and most alluring interface ever seen, we wallow in an agency-model that delivers a diet of over-designed cartoons, stock images, animation, badges, gamification and every other damn distraction we can think of. ‘Keep It Simple Stupid’ has been replaced by ‘Keep It Stupid Stupid’. Google focus on back-end functionality to deliver a superb service, not front-end visuals. So should we.
Take the world’s most successful retailer, Amazon, with 44% of all online sales. They are obsessed with customer behaviour and simplification – not aesthetic design. Their website could be described as quite ugly, but it’s a masterpiece of cognitive simplification and the design of process and success, not aesthetic ‘look and feel’. They are successful because aesthetic design isn’t the point – selling and buying is the point. Similarly in learning – teaching and learning is the point. Like Google, they focus on back-end functionality to deliver a superb service, and do not rely on front-end visuals.
Social media
One could hardly describe Facebook and Twitter as relying on their designed interface or images for success. There are no Facebook or twitter images, there is no animation, only a core, scrolling timeline that draws you in and a simple interface that gets you typing stuff in. They understand that the goal is interaction, not spoon feeding, that the software behind the skin is where the real power lies. They understand that less is more.
Successful learning design
So how should we design for success in learning? First up, we need to focus on the outcome – successful retention and recall. This is our equivalent of Google's ‘finding the right thing' so that we click on it’ or Amazon’s ‘offering us the right thing so we buy it’ and Facebook/Twitter’s ‘interaction with others’. This comes down to a few simple principles:
1. Effortful design
Forget the graphic/text/graphic/text/MCQ model for one moment and think about the simple fact that the learner really does need to make the cognitive effort to learn. You have to make them think and act. The online learning industry is obsessed with the MCQ and their awkward cousins, the T/F, drag and drop and so on. Multiple choice questions are light touch, give the answer anyway and are poor on retention. That is because they are weak in terms of effort. You are not making the learner recall the answer from their own brains, rather, they are choosing from a list. It's an act of recognition. These interactions bear no similarity to how people actually use what they learn in real life. You have to know stuff, recall stuff, not pick stuff from lists or drag words from one place to another. If you don’t you’re designing for forgetting. So move to open input.
2. Simplicity
Google, Amazon, Twitter, Facebook, Netflix and every other online service, allows you to scroll down the page. They have largely abandoned the online learning, fixed-page model. Most online learning vendors have scrolling on their own websites but when it comes to learning design they default back to some old-school, fixed-page turning model. Sure you need to chunk material down but electronic page turning through coffee-book designed pages, is not the answer. No need to be flashy, Flash died for a good reason. You need to cut things down, get rid of those extraneous graphics – those stock photos of people in offices, looking at computer screens, managers smiling inanely at each other., patronising cartoons.... You also need to cut the text until it bleeds, then cut it some more. A good editor is of more use than a graphics designer. Forget those dull learning objectives at the start of your course, all of that Michelin-man padding. Sure, adhere to some simple rules on branding, through logo, palette and font – that usually means pre-defined colours but don’t get fixated on superfluous elements that distract. Your goal is learning and retention not aesthetic pleasure.
3. Get smart
Stuck in a flat HTML world where all the effort goes into page design and a flashy CSS, the online learning world hasn’t learnt from Google, Amazon, Facebook, Twitter and Netflix. AI is the new UI. As all the effort goes into the surface skin, there is no smart delivery behind the front-end. Google is pure AI, Amazon’s huge AI platform delivers what you see with subtle recommendations based on your personal behaviour and the behaviour of others. Social media is mediated by AI as is Netflix, which is why it has conquered the globe in the entertainment industry. Yet in online learning we are stuck with flat pages of HTML, with a few branches. Look at AI, that is now the real world.

We are in this pickle because we do not pay enough attention to learning theory. Anyone can say ‘that looks nice’ few can say ‘that’s great learning’ and justify their claim. What to do? Let’s get smart by using smart, behind the scenes software to drive the delivery of online learning. Let’s be honest and say that what we had was OK for that time but it’s time to move on. Let’s drop the idea that it’s all about ‘design’ and focus on functionality and leaning outcomes – what we actually retain and recall. Let’s stop being a nation of online shop, window-dressers and focus on learning, which is why we need newer tools and services, that can deliver effortful learning and work to principles of cognition that lead to learning not just looking.
If you're interested in this direction contact us on WildFire - the world's first AI-driven content creation tool. Or try a different approach.... adios....


shackletonjones said...

Donald, we’ve both been at this too long. The research is only as good as the theory, and today we don’t have a learning theory worth the paper it is written on. Without a solid basis for their decisions people understandably go on looks alone. There’s a pitiful range of trial & error and attempts to borrow from successful models (such as Google and iPhone apps) but these inevitably fail for the simple reason that, in the absence of understanding the learner, the approaches are predictably grossly misapplied (e.g. attempting to dump corporate videos on learners in a grotesque parody of YouTube).

So for the last 15 years or so I have been trying to show set out the basis for thinking and working about learning, and ignore all the noise in the meantime. If someone can’t demonstrate even a rudimentary understanding of this, I can save myself the horror of reviewing their book/blog/platform/app.

To this end, I should point something out: the successful end is not “successful retention and recall”. Google cares not a jot about ‘successful retention and recall’ – it’s just helpful. Though I appreciate this is not the comparison you wanted to draw, it is perhaps more apt than you realise: we will fail as a profession on this point – getting people to memorise stuff doesn’t shift performance – and our job is the latter. That education has focused disproportionately on rote learning will likely be its downfall too.

On the philosophical point (about whether learning is memorisation or not), it is clear to me that we NEVER retain information. We only ever reconstruct it from our reactions. Whilst that reconstruction may turn out to be substantially accurate, does not alter the fact that this is confabulation, not recall.

I know you’re optimistic about the AI direction, but consider this: my SatNav doesn’t give me ‘timely learning nuggets’, it just tells me what to do. There’s a big difference.

Donald Clark said...

Thanks Nick. Always value your feedback. I was, as you noticed, using the Google comparison on the UX front, not the learning front. I do diasgree on the goal of retention and recall. Sure recall is a reconstrucitve act - with the downside of fallibility, even false memories, that doesn't mean it is not useful - the slip from reconstruction to confabulation is a slip too far here. I'm heavily involved in Medical training at the moment and it really does matter there and in many other areas. I totally agree that the focus on rote learning is overplayed that does not mean it is not useful. I really do want my kids t know their times tables but don't feel it is necessary beyond a certain level, in other words, there's a line to be drawn and it's often drawn in the wrong place.
On AI, the Satnav helps you get to your destination, as an aid. Similarly, I've seen AI help get learners to their goal, Google is a clear example if you're doing research (AI aided research) as have others. Important not to write this off without doing the evidence-based work... some way to go but it's more than promising. Adaptive learning has great promise, AI aids have great promise - chattiest look interesting... skills-based AI-led tutoring... lots to explore here...

Mark said...


Pointed and clear writing as always and I think you'd agree, not a new chorus that a number of us have been singing for years apparently to deaf ears. The L&D field though isn't populated by ill-intentioned folk seeking to commit bad acts so I have to ask myself why has this model persisted? I'd argue that what we don't talk about enough is the ecosystem that L&D lives in. There are curricula of instructional design that teach a certain way. There are corporate interviewers who are looking for certain things in a portfolio. There are RFPs that demand adherence to certain requirements. There are institutional budgets that demand certain measurements. None of these vectors gives one whit about learning theory per se. In fact, if they do, its often in spite of their own environments. As many times as we mention Google, we should also mention Lycos, Hotbot, Alta Vista, Dogpile, etc and ask why did Google win? I'd argue it mastered its business model. Now though we're faced with not arguing about why evidence-based learning theory isn't more widely applied but we're up against entire economic systems that reward behavior divorced from any consideration of design or theory. Let's disrupt those systems. Cheers, Mark

Elisabeth Siegel said...

Thank you Donald for the article. I greatly enjoy reading your thought-provoking blog. This article reminds me of the priorities in instructional design and in the industry in general. I completely agree with you: If we have to decide between great visual design and effective didactic methods, the priority should be the latter. Yet, my vision is that we do not necessarily have to make a choice. Possibly, we need both. In my mind, good design means that “form follows function”, i.e. the visual design supports the didactic concept. However, in order for the visual to support the didactic you need a solid understanding of how people actually learn. This is what makes the role of instructional designers so demanding: You need to be a mix of analyst (processes & performance needs, business goals, …), didactics expert, copywriter, UI designer, digital native … . And this is just a selection, not a comprehensive list. Are there so many people who combine all of these competencies?
I also believe that many people have a need for beauty, for the aesthetic (John Muir said: “Everybody needs beauty as well as bread”). Beauty can be a means to attract attention & attention is one pre-requisite for learning. I am completely against flashy design just to make “content” look fashionable. However, human beings are also driven by emotions. If we can make learning a pleasurable experience through delightful visual AND sound instructional design that would be it, in my opinion.
I also agree with Nick: If we can improve performance without any learning intervention, the better!

Donald Clark said...

Thanks Elizabeth. Broady agree. The question is whether 'beauty' is a necessary condition for success in learning. When I read a book, I don;t expect lots of beautiful graphics, just a good font and black on white text - I wouldn't buty a book that had any other colour combination. The problem is to confuse beauty with 'graphics' and 'images'... but generally agree. A bigger problem is that too much of the budget goes on extraneous graphics and design and not much thought into learning... hence the casual acceptance of MCQs and 'click on Peter to see what he thinks... followed by a speech bubble.... cartoons... etc etc