Simple presenter to-camera videos seem easy to produce but, perhaps because they are so basic, it is tempting for people to want to sex them up with fancy graphics and text. Before you go there, take a look at the research.
Video is a great tool for sharing visual and audio streams of information and generally we have no problem processing these two streams simultaneously, but if we add in text as well there is a problem.
Don’t talk and read
Mayer (2003) and others have shown that the addition of text to a video presentation will overburden the viewer’s cognitive load. Captions, text or scripts on the screen while a presenter is talking actually inhibit learning. The one exception to this is language learning, for obvious reasons.
The research also shows that complex backgrounds and music beds will similarly overload cognition and inhibit comprehension and learning (Brahme 2016).
Learners perform less well with video when:
- In addition to animation and a narrator’s voice, there are subtitles or regular text summaries of the content
- The video shows interesting but essentially irrelevant information
A good starting point for learning video is to scrap everything that is not strictly relevant. Trying to enhance a video with fun images and text is usually counterproductive for learning unless the images and text are “to the point”.
Chunking is a well-known technique for avoiding cognitive overload and increasing retention of learning content. Long form video is generally better broken into smaller segments. Four segments of five minutes will work much more effectively than one long video of 20 minutes. This has been proven empirically.
Ideally, chunking will be combined with visual rests, where learners can reflect or engage in activities that encourage cognitive engagement with what has just been seen. Active and effortful learning activities will both enhance learning by reducing cognitive load and increase reinforcement, retention and recall.
My pragmatic response to this research is that we need to be informed by it, but it’s not possible or desirable to 100% stick to it. We can use a little text in our presenter-led videos, but it must be used very sparingly. In addition, the words chosen should be words from the presenter’s script in order that we reduce cognitive dissonance. So, if the presenter says “It’s most important that you wear protective equipment on site at all times”, a keyword text should not say “PPE”, but “Protective Equipment” to exactly match the voice.
Better still, use icons or clear imagery to illustrate points rather than text.
While subtitles are an important tool for accessibility, it’s unfortunate that subtitles also increase cognitive load. So, where possible, design the video delivery so that the subtitles can be turned off and on.
Most of this information is gleaned from Donald Clark’s Learning Experience Design and Urban Myths about Learning and Education by De Bruyckere, Kirschner and Hulshof, although I did take a peek at some of the original research.