Archive for Instructional Design

Thursday, September 25th, 2014

Materials in motion: Exploring the use of animation in learning

By: Aaron J. Dewald, Originally posted at the Center for Innovation in Legal Education blog

A year or so ago, a discussion on Teknoids had started about what defines a video. It was inferred that something needed to move to constitute a video. Do narrated slide shows from Keynote, PowerPoint, etcetera constitute videos?

Legal education is going through some interesting changes at the moment. Much of it is related to distance education: online, blended, etc. Those who are starting to dabble in it are exploring a whole host of options for delivering this type of education: talking headnarrated presentationshand drawing on white board, etc. Coming from a learning science background, I’ve read an array of research out there that talks about this very thing. They all ask the question: What kinds of digital materials should we be using? What is the impact of a “video” on learning and comprehension? Can including animation (things that move) improve learning outcomes by students?

Let’s set an operational definition of the of the word “video.” In the learning science literature, you typically see references to multimediavisualization, or visual representation instead of the word video.  For purposes of this blog post, I’ll refer to it as a visualization.  Further, visualizations can be static or dynamic. This allows us to refer to the content, and not the vehicle that contains the data.  I mean, technically, anything in a .mov or .mp4 format is a video…. because it’s a video format, right?


Looking to the Theory

Let’s start off with a definition of the word “visualization.”  As I’m sure you can imagine, a visualization for these purposes is something someone could see – something visible.  In the literature, visualizations are often separated into two different types: static and dynamic.  A static visualization is one that does not have a temporal component.  Nothing changes over time… it’s just… there. Like a photo or a drawing.  A dynamic visualization is one that does include a temporal aspect.  This is a key distinguisher when it comes to making decisions about the type of visualization you’d use in your class.

Let’s chat for a brief moment about static and dynamic visualizations. The materials that are used in instructional environments need to be carefully chosen to accomplish certain learning goals.  The proper materials at the proper times can build learners up to the relevant level of knowledge they need. In the literature, this is called scaffolding. Now, with the nature of the Internet, there are so. many. materials. to. choose. from. The problem is, is that you really need to think long and hard about what you pick to use in your class, presentation, or other environment where learning occurs. The appropriateness and effectiveness of static or dynamic visualizations depends a lot on what it is you’re teaching and who you’re teaching it to?  Novices demand simplicity and concreteness (Goldstone & Son, 2005).  Experts do better with theoretical constructs.  Novices can get hung up on seductive details that are irrelevant to what is learned (Clark & Mayer, 2011).  Experts do a better job identifying what is important and focusing on that.

When I first started my PhD program, I thought that dynamic was always a better choice.  I mean, TV is dynamic. YouTube has tons and tons of dynamic content. This idea that dynamic is always a better choice almost seems supported by the literature.  Hoffler and Leutner (2007) did a meta-analysis of studies that directly compared static and dynamic visualizations.  As an aside, a meta-analysis is one where you don’t necessarily conduct your own experiment – rather you aggregate many experiments and analyze those pieces as a whole. They examined 71 studies that directly compared static and dynamic visualizations. They reported a significant effect of dynamic over static. Specifically, they found this effect when the visualization contained procedural information.  Sweet… so blog over, use dynamic… right?

Not quite. Other research done wasn’t so certain about this.  Specifically, Tversky, Morrison, and Betrancourt (2002) suggested that dynamic visualizations aren’t all that Hoffler and Leutner cracked them up to be. They suggested that dynamic visualizations are only useful when the concept being portrayed has a necessary and natural temporal aspect. Basically, if the movement is necessary to depicting the “thing” being learned, then it should move. When Hoffler and Leutner did their meta-analysis, they examined an awful lot of studies that examined the effect of dynamic visualizations in teaching learners procedural information: tying knots, assembling machine guns, physics, hand-bandaging, circulatory system, dribbling a soccer ball, etc.

Tversky et al. offer two principles that can help decide whether or not dynamic visualizations should be used, the Congruence Principle and the Apprehension Principle. The congruence principle is rather simple. Essentially the visualization should match up to what it’s representing. This is specifically aimed at those that try to be clever and teach something in a way that’s completely unrelated. The apprehension principle is equally understandable: keep it simple.

Additionally, they think that learning from dynamic visualizations can actually be bad for some learners. Sometimes, the animations are too complex, contain irrelevant movements, or are too quick for students to learn from.  Basically, they aren’t always designed very well.  These inappropriate designs can impact the cognitive processes the learners use.  For example, information in a dynamic visualization is transitory. Things move on screen, things move off screen.  Pictures and other images are here, then gone.  Narration is spoken… then never heard again.  This transitory information makes it difficult for some learners to attend to.  Further, asking the learners to integrate all of this with other transitory information as well as the information they already know (prior knowledge) might result in an excessive cognitive burden and divides their attention over the length of the visualization.

Let’s extend this a step further. Sometimes, designers or teachers will create learning materials that are dynamic for no other reason than, “because they can.” Prezi is a good example of this. Done right, Prezi works well for weaving a non-linear story. Most of the time, it’s good for making people sick and fragmenting their train of thought. Theoretically, this was studied by Lowe (1999). Lowe tested learners’ ability to learn from complex weather animations. They were asked to study from an animated weather map that taught certain atmospheric conditions.  Their post-test examined their ability to recreate the phenomena they saw in the study materials.  It was found that the learners focused on the wrong things. They spent more time looking at the perceptually obvious details as opposed to the perceptually relevant ones. The flashy things distracted the learners from learning. Do you want your viewers to focus on how cool the movements are or the information in your presentation?

The learners in Lowe’s case all had one thing in common. They were novices.  These were learners with a limited base of knowledge in the domain. This brings up a valid point. When you choose materials to use in class, be sure you understand the prior knowledge of your learners.  Many 1L’s come to law school with a very, very limited prior knowledge in any legal domain. Because of this, we should be using simple graphics and images.  We should be using relevant animations (also simple) if necessary.  Just because it looks simple, doesn’t mean it’s “childish” or “non-professional.”  Remember, research has been done on this, so if someone calls you out, point to the literature! 🙂

What have we learned so far?

  1. Learners don’t always learn best from movement. Static might be just fine, so long as there isn’t relevant temporal information involved in the topic.
  2. Dynamic visualizations can be detrimental to some learners, especially if the visualization moves too fast or asks the learner to integrate too much. Comprehension is a process that takes time.
  3. Know your learner. Novices might struggle a little more with dynamic visualizations than more expert students.

Implications for learners and suggestions for legal education

What does this mean for our students?  In today’s learning environment pushing blended learning, this should be a key consideration.

Many faculty and instructional designers want to get really fancy with their blended learning materials.  They realize that forty minutes of a faculty member lecturing to the camera isn’t such a good way to teach online. In reaction, some will go all out, finding ways to use Flash or After Effects to create elaborate animations or cartoons. Some might feel “inadequate” that all they’re doing is narrating PowerPoint.  What we’ve just explored should bring your concerns to rest. There are some anecdotal reports that they don’t need to be flashy.  They can be casual. They can be simple.  They don’t need to entertain. As long as you keep the principles we talked about in mind, you’ll do fine!

To wrap it up, law isn’t necessarily filled with lessons where time is a necessary and natural component of the topic. Much of law is interpretation and application on a foundation of relevant facts, rules, statutes, and more.  Nothing in there necessarily lends itself to being displayed dynamically. This works out in favor of law schools that are seeing budgets slashed and technology departments shrink. There’s no need to have elaborate equipment or software to create visualizations that are used in class, online, or other places. It does suggest the need for individuals that understand multimedia literature, to some extent, but I feel law schools will need instructional design or educational psychology help in the near future… so be one of the first and hire an instructional designer.

Thursday, October 10th, 2013

Improving Presentations (or videos, or other multimedia) with Learning Science

Note: This blog post was derived from a presentation I gave at the New York Law School. I was invited by Doni Gewirtzman and Kris Franklin to speak about the impact of learning science on the creation of presentations. I realized there are many nuances to the use of presentations. Some lecture with them, some don’t. Some use only a handful of slides, some use a deck of sixty. These are very general and basic principles that I think cover the widest range of situations. However, there are absolutely many ways in which this can be interpreted and implemented. For example, if we’re walking through language from a contract, then yes… it’s absolutely ok to have a mess of text on the screen. We can do simple things like text fading and using boldface fonts to control attention.

That said, I hope you find this post useful. Many of the concepts are at their very basic, but as we move forward with creating materials for a blended or flipped classroom, consideration to how we design these presentations/multimedia will be necessary to improve the efficiency of the teaching and learning process.

Finally, I apologize for the length of this post.  There’s a lot of information to communicate. I’ve provided three multimedia videos that help illustrate some of the points, as well as break up the information into three separate chunks. The content in the videos is the same as the content in the text, so you get to choose which information intake method is most effective for you! 

Plan of Action
This post will cover three primary topics.

First, we’re going to talk a quick bit about what we ultimately want for our learners. I’ll introduce two key terms from the learning science literature, then offer explanations for what they really mean.

Second, we’re going to talk about the impact of these concepts on the presentations we use or the multimedia we expose to students. I’m going to be talking specifically in the context of effective presentations, but this is really where the rubber hits the road for other types of learning like blended or online courses.

Third and finally, based on the theories and principles we’ll have discussed, I’ll offer three practical examples of how this information can improve your presentations and then ultimately benefit your students in their comprehension of learning materials.

A Quick Introduction to Learning Science

1. What is Learning Science

As mentioned, I’m a student of the learning sciences. At it’s heart, learning science is:

“An interdisciplinary field that works to further scientific understanding of learning as well as to engage in the design and implementation of learning innovations, and improvement of instructional methodologies.” Wikipedia

So based on this definition we can see that much of what is currently going on around us in education is starting to base itself on scientific understanding of how people learn, then reacting to those results through improving ways in which we currently instruct AND innovating new ways to promote learning. Blended learning and online learning are examples of these innovations.

2. What can it tell us?
Learning science can tell us what we should do for our students, and back it with science. Obviously, we want our students to learn the most they can. We want them to engage with this information, to process it, to understand it, to reuse it, to apply it. Learning science can advise us on how to best accomplish these learning objectives. The first idea will inform us on what levels of understanding our students have, telling us how much they know, and how they know it. It’ll also help us adjust the way we teach to make sure everyone has an opportunity to learn the same amount. It’s the idea of Depth of Comprehension.

3. Depth of Comprehension
In terms of human evolution, depth of comprehension is a relatively new topic. In 1988, a professor by the name of Walter Kintsch (who just so happens to be my advisor’s advisor) offered an explanation of learning called the Construction-Integration principle. For this post’s purposes… what Construction-Integration is isn’t as important as what this actually means to us. The CI principle tells us that there are three levels of comprehension that learners can possess: surface, textbase, and situation model. Let’s take a quick stroll through these.

A surface level of comprehension is one in which the information is learned absolutely. Think of memorizing a poem. If there were one in a language you didn’t know, you could probably “learn” it based on sounds and intonation… but you wouldn’t have a clue as to what it means. Generally speaking, this rarely happens in learning, so we don’t talk much about it in my field.

The textbase level of comprehension is one that is much more prevalent in learning. For ease, we’ll call this the shallow level of comprehension. A shallow level of comprehension occurs when a learner “learns” just enough to paraphrase the main ideas back to you, when they learn it just enough to pick out items from a multiple choice list. A shallow level of comprehension occurs primarily when students cram for exams. Hoping to get the gist just long enough to take an exam and do relatively well in it.

As you might imagine, this level of comprehension isn’t the best level of comprehension. Why?

A shallow level of comprehension can result in inert, unconnected knowledge. It’s inflexible and unusable in novel situations… situations that many learners encounter in legal education. Further, because it has such a transient existence in a learner’s mind, it’s often out just as quickly as it gets in, leaving it outside of the long-term memory of the learner.

What we want our students to obtain is a situation model level of comprehension. We’ll call this one deep comprehension. A deep level of comprehension is constructed when the learners truly absorb what is being taught. They integrate the information with their prior knowledge. They connect what they’re learning to what they already know. They have a well-developed structure of knowledge, which is a hallmark of deep comprehension. In turn, they’re able to apply this new knowledge to novel situations. It’s flexible, so they can apply it to new problems they encounter, inside and outside of the domain in which it was originally learned. And because it’s integrated with their prior knowledge, it’s much more stable and permanent than when it’s a shallow level.

Unfortunately, as you might imagine, deep comprehension is also the most difficult level of comprehension to achieve in the classroom. There are some things we can do, however, that help us give students an opportunity to develop the deep level of comprehension we hope.

Two Theories that Inform Multimedia Design

1. Introduction to Multimedia Theory
Fortunately, there is one way in which we can immediately impact a learner’s ability to develop a deep level of comprehension – through our presentations and other multimedia. They way we structure our presentations can have a profound impact on the level of comprehension our learners can achieve. To better understand this, we’ll turn to two key ideas in multimedia design: the dual-coding theory and the other is the cognitive theory of multimedia learning.

2. The Dual-Coding Theory
The dual-coding theory was developed by Allan Paivio in 1971. At its essence, the dual-coding theory explains that when you provide learners with two ways to learn or encode information, they can understand it more deeply than if you only give them one.

Paivio offers the reasoning for this is because different kinds of information are processed differently and along distinct channels in the human mind.

Verbal information goes in through the ears and is processed as verbal information, visual information is taken in through the eyes and processed as visual information. The brain can then “smush” these together and allow them to build on themselves and provide a better encoding of the to be learned information.

So, put into practice, it looks like this. I can give you the word “dog.”

I can give you a picture of a “dog.”

But, as the dual coding theory goes, if I provide you with a picture of a dog as well as the word “dog,” the fact you’re seeing both allows you to encode this information more thoroughly than receiving one alone.

When the verbal and visual contents are complementary and overlap, there can be multiple retrieval cues, which enhance recall and in some cases a deeper comprehension of the information.

One thing to understand, though, is that learners do not have an infinite amount of bandwidth. Humans can only process a finite amount of information in a channel at a time, and they make sense of the learning materials by actively creating mental representations. If we overflow their channels, we run the risk of limited comprehension by our students. The more load we put on them, the more likely it is that shallow levels of comprehension will be attained.

3. Cognitive Theory of Multimedia Learning

A professor from UC Santa Barbara by the name of Richard Mayer came along and went one step further and offered the Cognitive Theory of Multimedia Learning. This theory was proposed in 2001. Its basic idea builds on what Professor Paivio found with the dual-coding theory. Words and pictures are superior to words alone – with a few caveats.

The Cognitive Theory of Multimedia Learning consists of 10-12 individual principles. These principles have held up fairly well over the last decade but many of them have been tweaked as more and more people turn to learning with multimedia, such as online classes, hybrid and flipped learning, and so on. I’ve chosen three of the most relevant to talk about: multimedia principle, coherence principle, and the redundancy principle.

If you’re really interested in the other principles, which are no less important, you should buy his book Multimedia Learning (Cambridge Press, 2001). There’s likely an updated version – I know he’s done more work refining the principles since the blended learning revolution is upon us.

a. The Multimedia Principle
The multimedia principle bridges us from the dual-coding theory to the multimedia principles.

Basically, it states that learners learn better from words and pictures than words alone or pictures alone. Recall this is something that Paivio also advocated for. Mayer studied how students engage with different types of learning materials and found that students’ comprehension was more shallow when they learned with text alone – they weren’t engaging in deeper processes that allowed them to connect the new information with what they already know.

So, you might imagine this comes into play when we have PowerPoint slides that are nothing but text… as far as the eye can see. Our learners aren’t really learning much from that. There’s so much text on the screen, the best they can do is try to “get the gist” of the slide and engage in shallow learning tactics to absorb it all.

b. The Redundancy Principle

The multimedia principle leads directly into the next principle we’ll talk about.

The redundancy principle states that learners learn best from pictures and narration than from pictures, narration and on-screen text – especially when that on-screen text is the same as what’s being spoken to them. The redundancy principle offers that learners don’t learn well when they hear and see the same verbal message during a presentation. Many people will use their PowerPoint presentation as a script, and read from the slide, offering very little new information that isn’t already on there.

This brings up an interesting point that involves the dual-coding theory. Remember how we said that images are processed through the visual channel and sounds are processed through the verbal channel? Well, our channels are not infinite and have a pretty limited capacity. When visual stimuli like pictures are presented, the visual channel picks those up, no problem. When spoken narration is heard, those are picked up by the verbal channel. Text, however, is weird.

Text, you’d think, is picked up by visual channel, but it’s not. Text actually imposes on the verbal channel. You can test this out by reading something and having someone talk to you. It’s very hard, if not impossible, to take both information in. So, if you’re reading your slides and your learners are reading your slides, you’re consuming a lot of their brain power. This is brain power they could use to form deeper connections with what you’re talking about, formulate questions, or make inferences. If we overload one of the channels, we run the risk of shallower processes.

The redundancy principle basically asks to minimize the amount of text that you put on slides. Reading is very instinctive, and people can’t help it. if you have a mess of text on a slide learners are going to read it, if you’re talking, you’re adding to their load. Learners can’t process both at the same time. If you’re going to say it, don’t put it on the slide. Just use a word, or a keyword, then speak the rest.

c. The Coherence Principle
Finally, we’re going to chat about the last principle, the coherence principle. It states that learners learn best when extraneous pictures, words, sounds and other such things are excluded rather than included. Essentially, less is more.

Things that move, sparkle, or shake can confuse novice learners. They may be distracted by the movement. They may try to figure out what it means and spend some of their processing power on something that isn’t relevant to the topic. This also goes for images that have nothing to do with the topic. When you select images, they should be relevant to the topic to help them understand or to illustrate a point.

So, that fancy animation you found online? Ditch it. The sparkly transitions between slides? You don’t need them. Random clip art or pictures that “looks cool.” Leave it out. Anything that doesn’t directly contribute to the learning, understanding, and comprehension of information should be left out. Use movement to draw attention to salient and important details, not to wow the audience.

d. Depth of Learning
The multimedia principles were studied for the depth of learning they provide learners, and through Mayer’s experiments, they found that adherence to these and the other principles can help in providing learners with a deep level of comprehension by making connections more concrete and allowing students to focus on what’s important and seeing how things may relate to one another. This is done by freeing up some of their cognitive capacity, allowing them to think more deeply about the topic. This is done by connecting words to images to show examples of topics or concepts.

While there’s no perfect solution to designing multimedia to always result in deeper comprehension, every little bit can help.

Practical Implementations

All the theory in the world isn’t worth much to you if it can’t be applied to your day-to-day operations. This is where the fun part of my job comes in. I’m going to talk about three ways in which the research we’ve just described can be applied to your presentations.

1. Less Text, More Visuals
This is the one you saw coming. Both the dual coding theory and the multimedia principles advocate for less text on the screen for your learners. This is sometimes hard to do. Many professors I work with use their presentations as a script… in fact, many people do this, but while it might make things easier for you, it also distracts from the learner. Large passages of text compel a learner to read.

When they’re reading, and listening, and watching, they’re overloaded with information and some will be lost. Since their verbal channel is being overloaded by reading and listening, there isn’t much bandwidth for deeper comprehension, they’re trying to read and write and listen. Typically their comprehension forms on a shallow level.

I offer the suggestion of using key words or short phrases as anchors for your talk, then articulating what they mean, use fewer bullet points on a slide. Rather than putting five or five bullet points on the screen and repeating them out loud, use three. The less text you can put on the screen, the easier it’s going to be on your learners. When selecting images, find ones that can help support your message.

2. Simple Graphics are Okay, Preferred in many Situations
When you’re assembling your presentation, don’t think that you need the fanciest, most detailed or complex images for your presentation.

Research has shown that overly complex imagery can distract learners from what is important. As an example, a lot of research has been done teaching things like the heart and circulatory system. When they use a “real” looking heart as opposed to a simplistic looking one, students didn’t perform as well. This is especially important when you’re dealing with novices and brings the coherence principle to the forefront. Novices can get stuck on the little details, like the veins and wrinkles in the heart.

As you think about which images to use for your presentations, they don’t need to be, and shouldn’t necessarily be, super detailed images. Silhouettes and stick figures work well. If you do need to show a complex picture, don’t put a bunch of them on the screen at the same time. Choose one that best illustrates your idea and stick with it.

3. Fancy Transitions and Animations Can Distract, not Enhance
I often get asked about transitions. Should I use them to “pep” up my presentation? Which should I use? Which are best? My answer is most always, none. These are known as seductive details.

When you’re creating a presentation, you want to make sure that you draw attention to two things. First, is you. You are the information source. You have information that you need to transfer to your students. Second, is your presentation. Your presentation supports what you say. Your presentation helps learners connect the dots, helps them find ways to infer and understand. You don’t want your learners to be distracted by the presentation.

Fancy transitions, movements, and animations will only help your learners understand if the movement helps them understand the content. So if you’re teaching combustion, then showing a moving engine will help learners deeply comprehend the concepts. If the animations or graphics aren’t necessary, don’t include them.

Now, sometimes I do use something that people may view as animation. Sometimes I’ll slide in text, pop up images, or something similar. This is done to control the amount of information a learner sees at any given time. If I have five pictures, or multiple bullets, I don’t want the learner to have to consume all of them at the same time, filling up their verbal channel. I will pop up what they need to see at any given point in time. Or, I’ll use it to draw attention to a point or concept.


As you can see, there are a lot of ways to sprinkle a little bit of learning science into your presentation. There are numerous concepts that branch out from this, such as how to select graphics themselves, what activities should I be doing in class, is there a “medium” level between textbase and situation model? All of these questions are good ones, and could be answered by diving into the research a little more and applying that to your specific and unique case. Either way, education has a lot to learn from learning science, and the quicker we are to understand, adapt, and implement those ideas, the better off we’ll be.

Tuesday, July 2nd, 2013

De-Aggregation Method (Jason Fiske, Guest Blogger)

Since you are now reading this I can assume that either I did not completely bore you out of your mind in my first blog, or you are ignorant to my first blog post and I now have a new opportunity to bore you. Can’t wait!

I thought I would now spend one blog talking about our research-based solutions that resulted in our current setup of our online program, and spend my next blog talking about new ideas I have for the future. We conducted a massive amount of internal research through targeted student surveys (this could be several blogs in itself!), and also pulled from numerous other sources.

There were two main things we found through our research and observations: 1) when our adjunct professors were left to their own devices there ended up being massive inconsistencies with the way different courses were taught, organized and delivered online; and 2) our adjuncts simply did not have the time to do everything that I would have liked to see in an online course by him or herself (it is physically impossible).

Our solution: what we internally call the “de-aggregation method” of course instruction. We have split up the functions of each course into its core components and have a different person specialized in that part handle it:

1) Instructional core: We follow the flipped classroom method that has been discussed in other blog posts, our adjuncts create these videos and they are placed in our “study guides” (see below), and they also teach live-online with students coming in having viewed the created videos. This is the primary function of the course, as well as creating the assessments.
2) Feedback/Study guide creation/discussion forum engagement: We have a “course manager” for each course who is a professor that gives timely feedback on all assessments (within 72 hours), runs active discussion forums (20% of overall grade for students), and also creates study guides. We have one study guide per week that students work through before each live session. Essentially it breaks down a topic into sub-parts and tells a student to read 5-10 pages, then asks a few questions about those pages and has the pre-recorded 8-10 minute video of the professor on that sub-topic for the week there as well. There are usually 7-8 sub-topics per week broken down in the study guide in this manner. Each study guide starts with a concept map that charts the entire course and where that week’s material falls within the grand scheme.
3) Online set-up of course in the LMS: We have a separate person entirely set up everything online. I have a team of federal-work-study JD students who help me with this. I train them on how all courses need to look and feel, and this has resulted in a consistency of format and delivery of information for all courses. Professors and course managers have very little power to change things themselves within the LMS, though they can always post announcements. Naturally, not all the professors liked this (though some were actually relieved) and I would not recommend this in all instances. However, we have seen a substantial improvement in satisfaction of course organization as a result.

If you would like more detailed discussion on any of these topics, check out my blog on online education .