Your Source for Learning
Technology, Strategy, and News
    [Forgot Password?]
ARTICLES      
RSS feed RSS feed

How Much Narration in eLearning? Our Lessons Learned

"We wanted to know the preferences of our employees so we conducted a survey. They almost unanimously said that 1) they do not want the entire course to be narrated, 2) they do not want text on the screen read to them word for word, and 3) about two-thirds of the employees want to be able to turn the narration on or off. "

How much audio narration should we use in our eLearning courses? Are we using the right narrators? How good does the quality have to be? Do we have the right equipment?

These are questions our three-person, internal eLearning team recently asked ourselves. We want to share our findings.

When our company first began using asynchronous eLearning about seven years ago, we used text and graphics only, no narration. Then we used audio sparingly, only enough to give a personal “voice” to characters in a workplace harassment prevention course.

Then about four years ago we received a mandate to develop a substantial amount of online compliance training. This would have been no problem, except that we had to ensure each employee met a prescribed minimum contact time. The only way we knew how to do this at the time was by using fully narrated screens, combined with deactivating each screen’s next button until “audio complete.”

This was not an ideal learning method, and full narration presented other challenges. We couldn’t afford paid talent, so we used employees. This added more challenges, such as finding suitable speakers, getting on their calendars, and matching audio quality of updates.

So recently we stepped back to examine how we wanted to use narration going forward.

How much narration?

We started by doing some informal research with industry cohorts and employees. How much narration should be used? Several eLearning gurus said they use little to no narration unless there is a bona fide instructional reason to do so such as augmenting online transaction processing with narration.

We found that providers of audio services and equipment favored a much higher use of audio than did our instructional design counterparts.

We wanted to know the preferences of our employees so we conducted a survey. They almost unanimously said that 1) they do not want the entire course to be narrated, 2) they do not want text on the screen read to them word for word, and 3) about two-thirds of the employees want to be able to turn the narration on or off.

Who should narrate?

Our industry sources all say that, if there is a narrator, the higher the quality of the narrator’s voice the better. We took this as a strong preference for professional voice talent.

However, our employees had a different view. Only 12% said they prefer professional voice talent. A full 85% said the voice only needs to sound good enough to get the point across without having to strain to understand it. Nearly 60% of our employees said “no preference” as long as the voice isn’t irritating to listen to. 40% prefer that the narrator be someone they recognize (i.e., a well-known manager, process owner, or SME). A surprising 9% said the narration could be computer-generated as long as it didn’t sound too robot-like. (We will report on our text-to-speech findings in our next article.)

How good does it need to be?

In addition to the narrators’ voice quality, what about the quality of the audio output itself? We considered upgrading the quality as much as we could. We looked at better microphones and became acquainted with Digital Audio Workstations, or DAWs. But then we realized that the cost and learning curve of these devices was not warranted because the audio will get compressed anyway, and because of feedback from our employees. We ended up simply getting a better microphone (about $150) and controlling the recording environment more.

Our new guidelines

Here are the guidelines we have adopted as a result of this study:

  1. [How much?] We will use audio only when instructionally necessary.

  2. [Control] We will make sure students have the ability to turn the sound on and off, and that they know how to do so.

  3. [Who?] We will continue to use in-house talent, but other than credits at the end, we will not identify the narrator unless his or her name or title is pertinent for the instruction, e.g., having the Compliance Officer introduce a compliance course. This will prevent having to re-narrate when someone changes position or leaves the company. We may audition to get more suitable voices.

  4. [Quality] We only need slightly a higher quality microphone along with a pop filter to raise our technical quality to the practical limit. We also identified a storage room that will double as our sound studio with the use of inexpensive draperies. This location should improve our ability to splice in updates without sounding noticeably different from the original.

  5. We will continue to have learners evaluate the use and quality of our narration and make adjustments accordingly.

References and web sites consulted

“Addicted to Audio?” Cathy Moore, Posted in Audio, Human interest, Instructional design by Cathy Moore on 13 November 2007.


Nuts and Bolts: Principles of Multimedia Learning,” Jane Bozarth, Learning Solutions Magazine (online), May 4, 2010.


Various posts on the ASTD E-Learning Discussion Board.



(101)
I appreciate this article

Comments

Login or subscribe to comment

Thanks for the updates on the use of speech in eLearning. I'm an ID, trainer, and eLearning developer, and I've also done quite a bit of professional voiceover for both eLearning and radio/tv commercials.

I agree that VO in eLearning can be annoying. And that you can get by without a"professional" voice. The real key is getting someone who has some energy and variety in their voice, and who can read well.

Like you, we've found that allowing learners to turn audio off is a must-have feature.

I look forward to your other findings.
Nice article. I agree with most points will offer that they come from an native language point of view.

If you are creating instructional content in English for individuals who don't primarily use the language, narrating the on-screen text can be helpful, and aligning the narration to the text is essential.
Thanks for the article. I agree this has become an issue.

I would have to add too that the quality to me includes the script. Sure people said they have "no preference" in voices when the script reads like a book. But if the script is well written, it ups the engagment level and retention level even higher.

Sarah Kesher
Sound80
If you can turn the audio off and not lose anything, then the audio isn't necessary in the first place and probably shouldn't be present. On the other hand, if the audio carries the weight of explanation--for example, if it is describing something illustrated with a still image or animation on-screen, then allowing learners to turn off the audio cuts the legs out from under the presentation.

The rule of thumb I try to use is this: use audio to describe pictures and animations; do not use audio on pages where text carries the weight of explanation. Recently I've taken to adding a little icon to each page: one icon for "audio present" and one for "no audio" so learners can see at a glance the audio status of any page.

I do agree with and try to adhere to the other principles you suggest: don't let the narrator read on-screen text out loud, and professional audio talent is not always necessary. With respect to professional voice talent, my experience is that such talent, because of how slick it is, can end up sounding institutional, rather than conversational. Since e-learning is an intimate, one-on-one experience between the learner and the course (personified by the narrator), an important consideration for me when choosing a voice is how authentic it sounds. Sometimes less professional voices are paradoxically better able to convey authenticity, because they sound more like real people.

-Ray
Thank you for sharing your findings. We have been trying to determine best practices for narration as well.

The pressing question for us at this time is how to incorporate narration into powerpoint classes without creating a file that is close to half a GB in size. The courses are university level and 50-60 minutes in length. Does anyone have any best practices around this issue?

I like the idea of using the storage space.
As a professional narrator, I don't disagree with your approach to finding out what works and what doesn't. However, I am a little dismayed that professional voice talent are not perceived as adding significant value to a project by this community, as most projects that noticeably stand out employ professional voice talent. I guess it depends on what you are hoping to achieve... mediocrity or outstanding achievement.

I would agree that some projects do not require a lot of narration, but for the projects that are narrative driven, why would you ever want someone from the office versus a professional voice? A professional voice knows how to enunciate, how to give words and phrasings true meaning. A professional narrator has a professional studio with soundproofing and a $500-$1000 microphone, as well as state of the art equipment. But most importantly, a professional narrator knows how to speak. I would ask you to put two presentations together, one with "joe" from the office, and one with a professional narrator. On the project with "joe" as the voice, listen for distractions like lip smacking, swallowing, big breaths, clicking, nasal congestion, echo, not to mention the fact that "joe" typically will "read" the script and in monotone, with no inflections or imagination whatsoever.


I would submit that if most of the people posting comments here agree with the article, you have most likely never used a really good professional narrator. There is no comparison.

Sincerely,

Keith Michaels
www.keithmichaelsvoiceimaging.com
I have to agree with Keith and Sarah. The quality of the script matters and not everyone knows how to write well for their audience. This talent is worth it and I'd recommend hiring (even hired guns) talent specifically tuned to writing in a specific style for a specific purpose for a specific audience. It's not easy. And everyone isn't ready or capable of it.

The second thing that bugs me is exclusively using internal talent for audio. This is more a question of efficiency and value than it is one of quality. Even though we might think quality of voice talent doesn't matter as much as other factors, this is one element of polish that is so easy to get right with a professional. This is one element of quality that can be a real suck on an unpracticed narrator's time. And if their rate is significantly higher than a hired gun narrator (known quality / fixed cost) you can end up paying MORE for less quality. This is crazy.

Everyone wants to think that given the tools they can do anything. I'd guess that's true. But there's a distinct difference between doing something and doing it well.
I was thrilled to read that an article was written to address audio in eLearning coursework. However, I found only a summary of opinion which unfortunately does not answer what is best for student learning.

The polling sought opinion from those making their living on eLearning, and neglected to include not only the opinion of student preference, but the most important deciding factor- the effect of audio in learning.

That being said, I did enjoy reading everyone's opinions on the subject, I would not advise making policy decisions on this opinion. The outcomes doe not align with research done with regard to comprehension and retention, nor does it not align with the student's preferences. The results of following this "study" with policy will yield a result that is not best for the student.

But, perhaps this article was written to support a belief that the delivery of an audio professional is not necessary? There is some merit to that, but also some folly in not utilizing professionals for your work.

The research on the subject of audio in learning vehicles is quite interesting, and there is no shortage of it. It is a subject of fascination for me, and since it is my business to know it, I research it and apply my knowledge and expertise in my audio.

Unfortunately, by the time audio scripts are sent, there is little change of any correction to be made. As is often the case, audio / voiceover is usually done in the final stages of course creation, and it is generally sent as a "do-as-you-are-told" scenario. It is treated as an after-thought.

When writing a novel, training guide, or a speech, the script matters. It should have consideration to the audience as the writing is different. Educators know this. They teach their students this important lesson, but it is not a consideration in most multi-media learning decisions that I find in my work.

But apart from the "scripting" of the content, there is the psychology and behavioral patterns of the students that oftentimes are either ignored, or lumped into a large category of "normal".

I do have the privilege of working with some quality coursework, and this past December I "fired" clients that didn't show Subject Matter Expertise in how they handled these important educational vehicles.

My main hesitation in placing my own children in an elearning environment is due to the unscientific approach in learning that the Internet has allowed. It brings the bad in with the good. I could not escape it this year with my own High School son as his school does not have an on-staff Spanish instructor. It is the only class where his performance can be considered POOR, and after reviewing the coursework and delivery, it is due to poor course design with not enough audio interaction.

Audio recording of every item on the screen has a benefit to the learner in certain subjects, and is more important in certain segments of your learning audience more than others. The obvious sight-impaired student will find the aided narration beneficial, of course. But, age, gender, developmental delays, at-risk children, and other factors play in whether audio should be included. Then the questions of how much, and how can be assessed as well.

There is so much research available and research supported with scientific study of the learner for those seriously interested in learning about it.

You are welcome to consult with me and I will explain what I know in relationship to your projects. You can launch your own research from there if you wish.

If you are an audio professional, feel free and contact me as well. We better our own industry by working together and if you know me, you know I love to share my knowledge!
I agree that audio can be overused, and using it sparingly is a good idea - when possible. However, our Fed government clients require that ALL slides be narrated (even if not word-for-word) for Section 508 compliance. In addition, all movies have to be close-captioned, and there has to be a text-only version of the course available. So "good eLearning practice" and research about the value of narration does not apply to clients that are required to comply with Section 508. Just my $.02 worth!
If your government clients think screens need to be narrated to be 508 compliant they are mistaken. Broad misinterpretation of these, and other, compliance rules is the cause of much waste and poor outcome.

On the other hand, if this is an assumption on your part there's another problem entirely.
Thanks for delving into such an important aspect of multimedia and eLearning.

I have to agree with the comment that if you can turn the narration off and get the same message, then what is the real value in it? There are few things as bad as having a narrator read text on screen verbatim.

In my experience narration can bring a lot to a multimedia eLearning course, when done right. It can also detract.

Two ways that I have seen narration detract are when the content is generally static and text-based, for example, a bulleted list of learning objectives. The other area is when it sounds amateur. Good voice over is hard to do, and generally speaking, you need a professional, (I know that many designers/departments do not have the resources for this, and you just have to do the best with what you have).

In my view, where narration works best is when it is used to drive the content as it develops on screen. For example, when illustrating a concept that might have multiple components to it, such as a life cycle, or where you want to emphasize relationships. Narration can be used to tell the "story" as it unfolds visually on screen. When the "story" is done, (narration complete), you are left with the visual artifact of the process or diagram that the learner can explore.

By the way, for 508, it is not necessary to narrate. In deference to non-sited users, you need to make sure that your content, (text-based and visual descriptions), is accessible to a screen reader. We do a lot of 508 eLearning that gets used throughout the Federal Government. I would be happy to talk with anyone, offline, about how best to achieve compliance and still have a good overall learner experience.
Very interesting example, was the target of the responses around compliance training?

In other words, I wonder sometimes if the employees did not want audio, simply because they felt they were forced to listen to it (for compliance training), and all they really wanted to do was "click next".

-Sean Bengry
I am a proponent of on demand narration. Some people prefer "hearing" rather than reading. Other people prefer reading. Sometimes the narration can help the learner focus.

I work with some clients who have difficulty reading and a narration link can be very beneficial.

I find the information on professional vs. amateur voice to be most interesting. I would like to see some research on this issue.
My personal experience is that, it is very hard to find an appropriate in-house talent that can read a script and make it look natural, not strident or boring. In some projects, especially when we are running out of time, I do the recordings myself even though I am not an English native speaker (my customers have never complaint about that …. Yet!)
Thank you for the article!
I'm currently reading "Efficiency in Learning: Evidence-Based Guidelines to Manage Cognitive Load" which is the latest from Ruth Clark and team.

It gives much in the way of research and evidence based advise on the "proper" use of audio. I stress "proper" because so many times it seems simply using audio improperly or without forethought has such negative results.
I appreciate the comments this article has generated. That was part of our goal, to explain the view of one team of practitioners and to stimulate the sharing of thoughts and ideas.

I just want to comment on one point expressed frequently in this discussion thread: the use of a professional narrator (and script writer). I think Ray articulated it well in his 4/25 comment where he distinguishes between sounding '...institutional, rather than conversational.' In the two companies in which I've been an eLearning developer, that is the sound we're after -- conversational -- regardless of the source -- and after hearing Ray say it that way, that's the real issue, not pro vs. amateur (with cost, real or perceived, added to the mix). We don't want narrators who sound like they're reading a script, nor do we want them to sound like they're talking to, not with, the audience. Also, we find that employees often prefer to identify with the speaker, and that often means they prefer to have a personal or at least organizational connection to the speaker as long as the voice delivery isn't distracting. It's impactful to hear about compliance from the Compliance Officer.

Finally, what we have not yet done are controlled studies of the actual effects of these options on learner outcomes in our environment. Hopefully we can do that one of these days.
Some people learn better through reading. Other people learn better through hearing. Providing narration that can be turned off covers both categories of learners.

I would suggest that some people might reject narration because of its low quality, not because of the presence of narration itself. No one wants to listen to someone who is obviously reading a script, possibly for the first time, and there's a continuous 60 Hz hum. Replace that with a more conversational delivery, by a pleasant voice, recorded with decent equipment, and suddenly no one seems to mind...

One thing that drives me crazy is when the narration doesn't follow the text on the screen. Do I read? Do I listen? Do I try to do both, and end up doing neither very well? Personally, I don't mind when narration follows text on the screen, if the text is basically bullet points and the narration fleshes them out.

Related Articles

Questionmark Perception is a popular assessment tool in the e-Learning community. Its three parts offer functionality for e-Learning designers and developers, and convenience for subject matter experts (SMEs) who assist with course development work. This article gives an overview of the Perception tool and a case study of user experience at the United States Department of Agriculture (USDA).
Some e-Learning projects are more challenging than others. This article traces such a project. The target audience was nursing instructors in higher education. Production involved making HD video in a hospital. The authors cover the project from inception to delivery, including content development, production processes, and lessons learned. If you want to make educational video, read this article!
From small-group presentations to stadium-filled sales seminars, Citrix connects people with a full range of video-enabled collaboration tools.