Why is it that sometimes animated characters make content more compelling and sometimes they seem, well, creepy? It seems like this can be the case, even for the same use, but just with different viewers. My friend and colleague, Professor Mark Salisbury, produces 2-3 minute segments on Knowledge Management for an NPR affiliate in Albuquerque, New Mexico among other locations. There are quite a few of these segments (and some supporting graphics) available on his web site (sample radio segments). I genuinely enjoy the segments for their content, as well as the recollections they inspire of good times and good fun working with Mark.

Recently he started using the Oddcast hosting service SitePal to make what I would call "pOddcasts" like this, The Knowledge Worker. I've heard about the research on animated characters and understand the need to more clearly personify the characters or vignettes in elearning-- I even encouraged it while at Macromedia (Animated Talking Characters for Elearning). However, in this case the technology just doesn't add value for me. I don't know if it is because I know Mark and have seen him speak, or because of my demographic (40+, non-gamer), or whatever.

Has anyone else run into this experience with virtual characters? I saw a presentation at the Learning 2005 conference that was titled, Can Virtual Characters Produce Real Results in Simulations? I really enjoyed the presentation by Tomas Ramirez and Greg Sapnar of Bristol-Myers Squibb. They cited some good fundamental research on use of virtual characters and included tools like a Virtual Character Decision Maker's Matrix job aid (from Night Kitchen Media) for selecting and developing characters. They referenced the work of Clark/Lyons and Reeves-Nass on virtual characters. This makes me wonder, was the issue for me that there already was a real character in my mind, so there is some sort of disonance when I see the virtual Mark? How can one ensure that virtual characters increase effectiveness and comfort for the learner?

Does anyone have other thoughts about this or pointers to relevant research?