Category Archives: Media selection

Optimal Length of Instructional Video

Someone recently pointed out this research to me. Prof Guo believes, based on the data collected from thousands of educational videos served on EdX, that six minutes is the optimal educational video length. People are willing to watch six minutes, but for every minute longer than that, the average number of minutes drops–drops not just as percentage, but in terms of the total number of minutes. If learners see that a video is 20 minutes long, they’ll quit well before six minutes (perhaps with good intentions of coming back later, but they don’t).

This is interesting for a lot of reasons, but one of them is that the governing body over learning for CPA firms recently gave the OK for firms to give formal credit for nano-learning. But to give credit, nano-learnings have to be at least 10 minutes long. EdX’s data* suggests that’s almost 70% too long!

*Of course, EdX’s data is not specific to adults learning in a professional setting. And naturally there are any number of factors in play that would make a three minute video seem interminable and a fifteen minute or even an hour long video fly by. But it’s still interesting stuff.


The Role of Technology in Improving K-12 Education

Like most people, I live in a school district that features an administration and staff that really care about students and about building their own skills as educators. They care about the community and about engaging with parents.

One initiative in the works in our district is an investment in technology to encourage educators to better use active problem-solving-based instruction.

At least, that’s what the initiative sounds like. In reality, I think it’s much better than that. I believe the effort is focused on supporting educators’ use of active problem-solving-based instruction and providing the education and tools, including technology tools, that can help.

The last two paragraphs sounds like they describe the same thing–more technology and more problem-solving-based instruction, but one leads with technology, the other with pedagogy. The distinction is critical because leading with technology makes adoption of technology the end goal.

Technology is not a goal in and of itself; it is an enabler for higher level goals. The higher level goals are the ones that need to be embraced. If they can be achieved without an investment in technology, great. If an investment in technology will advance the goal meaningfully, let’s do it. The critical piece is to stay focused on the end goals and not let achievement of the enabling goals alone be the criteria for success.

Compression: Live Training, Elearning, and Time on Task

I learned a term recently for something I’ve thought about in the past but didn’t know was a thing: compression. Compression means it will take learners longer to complete a live class than an equivalent elearning. In other words, if a four hour class is offered in both a live classroom version and a self-paced elearning version, participants will complete the self-paced version quicker with the same level of achievement.

Apparently–and I didn’t know this either–the rule of thumb for compression is 50%. A four hour live class can reasonably be turned into a two hour self-study. The persuasive utility of this, of course, is huge. What leader wouldn’t want their people in training only half as much time?

There are some enormous caveats. One is that time spent practicing cannot be compressed, so the more focused the training is on practicing skills, the less compression is possible. It also appears that the elearning has to be text-based in order to allow significant compression. People read faster than they talk. The elearning we produce at my firm are audio-based, so one wouldn’t expect very much compression.

That brings up the question, should our elearning be audio-based? We use audio for a number of reasons. The most important is that from a cognition perspective, using the audio channel for narration and the visual channel for complementary information (charts, organizational bullet points, tables, etc.) leads to better learning.

The question then morphs to: Is the increase in the quality of the learning worth the investment? Experienced professionals can take in new information very efficiently. If training is focused on them, it’s possible that the increase in learning may be immaterial compared to a time savings of 50%.

It would be interesting to offer the same elearning two ways (audio or text), randomize which one people get, and measure time on task, satisfaction, and exam performance across the two conditions.

Optimal Conference Size

RSM’s marquee learning events are its yearly internal conferences. They range in size from a few hundred attendees to over a thousand.

As the firm grows, the conferences have been growing. Inevitably, this will mean either larger venues or splitting up conferences into multiple smaller conferences.

The conferences are well run and well regarded. However, I had a partner in the firm express to me last year that while she enjoys the conferences, she misses the days when conferences were small enough that you knew or could get to know everyone.

That got me to thinking. Is there an optimal size for a conference? I’m aware of the Dunbar number, which suggests that humans struggle with group sizes above 150, but unsure of that number’s veracity or whether it even applies here. I need to do more research.

I think there is something to the notion that above a certain size, gatherings like conferences feel impersonal, easy to drift through and get lost in without connecting. But what size is that? And large, comprehensive gatherings also have advantages: a message of unity, a sense of size and accomplishment, the ability to feature the best and most powerful speakers, the ability to potentially connect with anyone across the firm. How much better does the user experience have to be to make up for these losses? Are conferences-within-the-conference possible/desirable? Do more interactive courses help mitigate some of the isolating effects of huge conferences by helping make connections? In the scheme of things, is any of this worth worrying about? So many questions.

The Learning Pyramid

A couple of months ago in a presentation I saw the following graphic.


“WOW,” I thought, “how have I managed to miss that research?”

Then: “Wait a minute, those are some pretty round numbers to be research-based.”

Here’s a good site talking about this model and how the pyramid might be a–probably directionally-correct though way oversimplified–model for making a point, the percentages on are just plain made up.

Classroom Salon

This is interesting. McGladrey has been talking about flipping classrooms but hasn’t done anything at any scale yet. What appeals to me about Classroom Salon is that it presents a way to envision the pre-classroom media portion of the class more than packaged, passive lecture, but rather as an interactive discussion. The way it works is that a document, or any artifact, I guess, is shared. Learners can read the document, highlight it, and comment on it, and the highlights and comments are viewable by all other learners in the class, who themselves can add highlights and comments, even threaded discussions. That sort of collaborative editing is not new–many programs can do it–but the novelty for me is in the use of document markup to replace or augment pre-recorded lectures as the prerequisite for entering a flipped classroom. I’m picturing instructors uploading important documents and embedding instructions like “highlight the biggest changes to today’s tax code,” for instance, and learners being able to see the most common highlights. Or the instructor seeding the conversation with specific questions about the reading, and then participating in the online discussions. Certainly in my experience, pre-reading assignments without accountability have low completion rates. This tool not only provides advisability tracking, but also makes reading a more active learning experience.

Three Media

At a professional development event a couple of weeks ago, a colleague made the following provocative assertion:

  • Elearning is for the “what”
  • Classroom is for “how”
  • Webcast is for review

That strikes me as at best a vast oversimplification for a number of reasons, but at least it has the (significant) advantage of being easy to explain to customers.

When I interview experienced instructional designers, I’ve started experimenting with asking them what they think of this model. Be ready for it if you interview with me!