When should elearning be narrated? I think we should rephrase the question as, “When is it a good idea to force all learners to go at the same pace?”
That’s what narrated material does. The pace of the narration controls the pace of the material. When you’re learning from narrated material, you can’t easily skim stuff you already know, or slow down and concentrate on the challenging parts, because the voice continues relentlessly at a pace that someone else established.
New studies suggest that learner control + text works better
According to recommendations in books like Elearning and the Science of Instruction, we shouldn’t narrate text that’s displayed on the screen. The redundancy interferes with learners’ ability to digest what they’re being fed. So is it okay to remove the text and use narration alone?
Apparently it is if you’re presenting very short science lessons that are based on graphics, which is what was done in the studies that are often cited. Most of the lessons were no more than 5 minutes long, and the learners couldn’t control the pacing. Those studies suggested that in those situations it’s better to use narration rather than text to explain a graphic.
But what happens if you use narration in material that takes a lot longer to learn, such as an hour? In one study, students who read silent text at their own pace finished more quickly and scored better on both retention and transfer tests than did students who used a narrated version of the materials.
“Our findings imply that the guideline to use spoken text can be restricted to situations in which time pressures are high and instructions are system-paced, based on the pace of the narration, and to situations in which there is a potential high cognitive load so that it is not easy to compensate by investing more mental effort. In all other cases, visual text seems the more sensible presentation mode, especially because it is cheaper to produce, easier to deliver, and in combination with learner-paced instructions even more effective in terms of transfer of learning.” (emphasis mine)
Another study supported this finding: Learner-paced, text-based materials got better results than versions that used narration or controlled the pacing. (Also check out this study if you need citations that show that learners learn best when given control of the pacing.)
Where are the complex graphics in our elearning?
Most of the corporate elearning I’ve seen has only incidental graphics—the ubiquitous smiling stock people, images of buildings, or scenes from a scenario. We rarely need to explain a complex graphic. So it seems even more likely that we can confidently display silent text on the screen and trust our learners to read at the pace that’s best for them.
And if you’re concerned about the accessibility of text in Flash, check out this blog, which covers Flash accessibility in detail.
What if our main goal is motivation?
The studies on audio focused on transfer of knowledge. As far as I know, they didn’t look at attitude change. If the main goal of your materials is to inspire learners or otherwise appeal to their emotions, a human voice could make a big difference—but not if it’s reading a dry script.
There’s a lot more on this topic at the Learning Circuits blog and in two previous posts on my blog: Addicted to Audio? and Should We Narrate On-Screen Text?
Image © iStockPhoto: nico_blue
The discussion is interesting but doesn’t cover two vital aspects. The first is chunk size. The bigger the chunk of information, the less the learner has control over the pace of delivery. Almost all the elearning I’ve seen has chunks that are far too big. Learners like small chunks (indeed, if the material is well written, they don’t even notice they’re small), while smes, managers, and all non ‘true’ learners detest them, so in they go, big, chunky and overwhelming.
The other aspect is the text/audio match. The range of sins in this area is enormous, and it’s not fair to extrapolate or generalise on the basis of a) narration that conflicts with the text and b) narration that is a verbatim reproduction of the text. Neither works, and neither does justice the use of audio.
And of course, no media mix can save dull, passive material. When it’s dull and passive, people look for scapegoats – often the audio, whereas it’s so often the learning design that’s the culprit.
Static text can be bland and mesmerizing, while a proper voiceover (from a pro, who can engage the listener and interpret your text) can make the connections necessary for most people to “get it” better. We had proof of this last year with the HearTheBill.org website, in which I and about 80 other voice actors, literally read all of the health care bills and put them up for streaming or download, making the lofty concepts and labyrinths of legalese more accessible to more people across the board. It is one reason why audiobooks are so popular, we take in as much if not more info thru our ears as thru our eyes, so the combination of both can be very successful.
Chris, thanks for your comment. I agree that professional voice talent can make a big difference, especially when the voice expresses emotion.
However, I want to point out that the HearTheBill page that I saw, as well as audio books, don’t display the text while they read it to you. They’re designed to work whether or not the listener is looking at a page of text at the same time. One of the benefits usually described by people offering audio files is that you can do something else while you listen. I’m a big fan of audio books and podcasts for that reason.
My main concern here is with the common and redundant pairing of text and audio, in which text is displayed while it’s read to us. As shown in the blog post, research suggests that the information would be easier to digest if it were offered in one mode only, not two redundant modes at once. In the case of HearTheBill and audio books, listeners have the *option* of reading along, and they have to seek out the text if they want to read it–it’s not forced on them.
Also, displaying text on the screen while simultaneously reading it to adult learners can be interpreted as patronizing, especially when it’s part of an elearning module that’s too easy overall, which is a very common problem. Obviously, the health care bills aren’t easy.
As for Middlemarch, mentioned in an earlier comment, I listened to it as an unabridged audiobook, which is how I “read” most fiction. I’m in no way anti-audio. I’m just tired of learning designers apparently thinking that adults must have their hands held and can’t be expected to read a screenful of text on their own.
Some comments have mentioned learning styles. People interested in an in-depth look at the usefulness of addressing learning styles in design might check out Learning styles: Worth our time?
I like the use of sound… IF I can turn it OFF when it is interfering with my personal pace or is redundant of words on the screen. However, when using Captivate’s motion features to guide a user through a sequence or process, it can be very helpful. To me, I think that it would be wrong to deny options to users. Give them the tools and control and see how far they fly!
Technology exists to control the speed of playback, and there is this thing call the volume control, or dare I say its simple enough to toggle narration and or audio off… I really don’t understand why there is so much debate regarding whether or not this feature or and other should be included in e-learning when it is not particularly difficult to give the learner some choice.
I don’t think the debate is about whether to add audio or not. It’s about choosing when to use it strategically. “Just because we can, we should” is a poor argument. Research indicates that attaching audio for everything on screen is a bad instructional practice, If reading everything on-screen is the choice, then you’ve excluded strategic and beneficial use of this medium for those that would choose to mute the audio. If the program is reading for me or going slower than my speed, I’ll turn down that volume or use the scrub bar to skip past a section (if available). If audio is added as a strategic element to extend visual media, I’ll miss it because I’ve turned off the main audio program. In my opinion, that’s a poor design choice.
There are great ways to use audio to support learning. There are less beneficial ways to use audio as well. As a rule of thumb, I’d say that folks SHOULD NOT exclude audio without first making a strategic assessment of it’s benefit. Audio is great when well employed. Audio is mediocre or worse when used to echo the onscreen text. This is compounding a potentially bad design decision with another potentially bad decision. Text is media. Use media with care.