Can Avatars Help Learning?

The main benefit of avatars is that they can help personalise a course, increasing its human qualities. For example, we have used photo/video avatars of people that the learners recognise and this has proved successful. It has also been shown that people find it easier to understand when they’re conversing with a person (although the jury is still out whether an avatar can achieve this).

However, it will be no surprise that if used incorrectly, avatars can be very distracting. (I saw this line referring to Microsoft’s paperclip avatar: “Looks like you’re planning to kill yourself. Office Assistant can help you write your suicide note…”) Other points to note are that guidance is fine, but learner control is also required, and avatars can’t (generally) take questions from learners (as you would expect in a normal conversation) so it’s best when the avatar interacts, by asking questions for example.

In summary, avatars are best used to enhance the message, not when they are the message – they should not turn courses in to instructor-led PowerPoint presentations. In another recent course we developed, the avatar was introduced in person at the start (and was a realistic, recognisable person), but we then only used their voice throughout (maybe popping up every now and again to summarise the key points) and this worked very well.

Regarding the use of audio, it is best when first and second person language is used for avatars. It has been shown that learners are more engaged and learn more by using "you" and "we" in the learning. Furthermore, according to Clark & Mayer, avatars should speak via audio narration, use a conversational tone, and have familiar human accents. On-screen text or machine-generated audio do not work as well.