The challenge of defining, explaining and evaluating immersion in the context of data worlds is tough. However, a simple example may lend insights. My wife Linda showed me a video of two kittens chasing a mouse, virtually on an iPad. Skip the ad and watch the first 60 seconds. Note what the brown cat starts doing at 50 seconds.
Cute! But let’s get serious… According to Milgram-Kishino virtuality continuum (for explanation), this is mixed reality where the cats are tricked into a lite immersive experience. They start with watching and then attempt to catch the virtual mouse. Later, the brown cat starts looking beyond the boundaries of the virtual space and even chewing it!
First, virtual immersion is a mind-trick, either self-imposed or inducted by others. In this case, it was by the cat owners for their entertainment. If we are to use immersion for good serious purposes, the subjects must willing suspend their firm gasp of hard reality and go with the flow. As IA designers, we probably should explicitly ask for their permission to inflict this mind-trick upon them, thereby ethically allowing them to make the decision to let go.
Second, there are all degrees of immersion, and the IA community needs to explore the effectiveness of the entire range. I tend to focus on TOTAL immersion where your mind is completely embedded in a complex virtual reality, solving some complex problem. However, this high degree of immersive is not best for everyone nor for every purpose. Lite immersion is more effective for general storytelling as evidenced by interactive 2D infographics by NY Times and others. What are the trade-offs and best practices?
Third, it would be useful to study seriously animal immersion. A search of YouTube would produce an abundance of examples. Anyone interested in collaborating on this project?