Meta Meta on Multimedia Design for Learning

Paul A. Kirschner & Mirjam Neelen

There’s a hierarchy of scientific evidence. At the bottom is evidence obtained through anecdotal evidence and expert opinions. At the top is evidence from randomized controlled trials, systematic review articles, and meta-reviews/meta-analyses.

Now there’s an article that might even rank higher on this ‘ladder’. Michael Noetel and seven colleagues conducted a meta-meta-analysis to determine whether certain multimedia design principles really held up. In their article ‘Multimedia design for learning: An overview of reviews with meta-meta-analysis’, they sought to (1) determine how multimedia can best be designed and (2) study to what extent both cognitive load theory[1] (CLT; John Sweller et al.) and the cognitive theory of multimedia learning[2] (CTML; Richard Mayer et al.) remained valid. They analysed data from 29 published systematic reviews, consisting of 1,189 studies with a total of 78,177 participants, all studying the effects of CLT and CTML on learning.

As the name suggests, multimedia uses different media, combining word (as text and/or speech) and image (both still and animated).

Keep in mind that a teacher making a PowerPoint presentation in class is ‘multimedia’. After all, the teacher speaks while showing images and/or words on the PowerPoint slides. Because multimedia lessons often contain a lot of information that needs to be processed simultaneously, it’s important to determine whether and how these multimedia keep the cognitive load under control. Derived from this, CTML focuses on how CLT can explain any variations in the effects of multimedia. At present, approximately 12 CTML effects have been defined such as how the following design strategies influence learning:

  • unnecessary distracting information (coherence effect),
  • signalling what is important (signalling effect), and
  • saying the exact the same thing in two different modalities (think of reading bullet points on your PowerPoint out loud, which is called the redundancy effect).

What they found was that many of the design principles were firmly supported. More generally, they found support for the assumptions that learners integrate information more effectively when presented through both image and sound, as long as care is taken to minimise the potential for cognitive overload.

They translated these findings into concrete directions for the teacher or instructional designer on how to improve the learning process:

  • Use relevant (and not fun or decorative!) images to support your story. This way, learners make the most of their visual and auditory systems (use them most efficiently in order to learn effectively). This also means that you should avoid words or images that are not directly relevant to what people need to learn. This is to make sure that they don’t waste mental effort digesting irrelevant content.
This might be an interesting image… however, if your lesson or eLearning is on ‘How to write a good resumé’, this image just distracts from what is relevant!
And this image might be funny, but it’s also not relevant for your resumé.
  • If you can’t find any relevant images, then only show keywords and not entire sentences or chunks of text (e.g., on slides). This way you can still ensure that learners can somehow use their visual and auditory systems. By the way, here’s a really good video on why (too many) bullet points DON’T WORK: Why bullet points don’t work – The Prezenter (H/T Mike Taylor – @tmiket).
  • Present words and pictures as close together in time and space as possible so that learners don’t have to use their working memory to connect them. For example, insert a previously presented slide into your presentation a second time as needed instead of saying “As you remember, the Krebs cycle… Just show it again!

And if you need to explain the parts of the brain, do it as shown in the image on the RIGHT (also see our blog on worked examples here). PS: The figure above also does this.


  • Help learners see (i.e. signal) what’s important by emphasising what you say, using a pointer, arrows, markers (bold), etc., so that can focus on what matters. Here’s a nice, short video on the signalling principle.
  • Don’t turn your slides into a Christmas tree. Think carefully about transitions, colours, fonts and font sizes, and choose a striking yet calm contrast colour to signal.
  • Only use animations and other moving images where helpful for learning. Four separate illustrations of the 4-stroke engine works better than a video, but an animation of a simple body movement (e.g., golf or football follow-through) might work better than a series of still images.
  • Use simple, personal language to help learners connect the new information with what they already know.
  • Break the lesson into meaningful chunks rather than one long, uninterrupted presentation. Learners should be given the opportunity to ‘consolidate’ their learning.
  • Only use a character/agent on the screen if they help draw attention to what’s important, and even then consider whether a simpler signal would be less distracting.

Conclusion: Multimedia works, but only when done right!



Noetel, M., Griffith, S., Delaney, O., Harris, N. R., Sanders, T., Parker, P., del Pozo Cruz, B., & Lonsdale, C. (2021). Multimedia design for learning: An overview of reviews with meta-meta-analysis. Review of Educational Research.

[1] Cognitive Load Theory (or CLT) is a theory which aims to understand how the cognitive load produced by learning tasks can impede our ability to process new information and to create long-term memories. See the YouTube video for a basic explanation Cognitive Load Theory 1 – An introduction – Bing video

[2] The cognitive theory of multimedia learning based on three main assumptions: there are two separate channels (auditory and visual) for processing information; there is limited channel capacity; and learning is an active process of filtering, selecting, organising, and integrating information (Cognitive Theory of Multimedia Learning (Mayer) – Learning Theories (