I read a study recently that looked at media spoilers. They wanted to know if giving away the ending of a movie/book/TV show ruined it for the person watching. What they found was really fascinating, but now that I have watched a few movies since reading the paper, it makes total sense.
What they expected was that knowing the ending would ruin the movie. We all probably think this intuitively. But what they found is that the movies that we like the best are the ones that follow enough of a typical formula that we can predict pretty closely how it will end, but the specifics about how it goes from the beginning to the end are not clear. So for example, we know that eventually the guy from the bad side of the tracks will fall in love with the girl from the good side of the tracks. But how will he do it? That makes it a good movie. Or in the case of a tragedy, we know that everything is going to go wrong eventually, we just don’t know quite how.
The two movies I saw this weekend illustrate this point. I didn’t enjoy either one very much, even though one was Jimmy Stewart and one was Hitchcock. I thought these were no-brainers as great movies. But in the Jimmy Stewart movie, I had no clue how it would end, and I found that I didn’t really enjoy the watching it unfold. I realized it was at least partly because I didn’t have a clue where it would end up, so I couldn't guess how it would get there and anticipate if I was right.
In the Hitchcock movie, I knew how it would end, but then it followed to formula so exactly that it was boring. What we really enjoy, is when we know the ending but not the process. The research I cited above found this among a large population (of college student volunteers) and it seems like it is true for me as well.
So why am I putting this on my HF blog? What does this have to do with system design and user experience (other than movie viewer experience)? Well, here is my take on it:
Remember several of my previous blogs about natural decision making. We used to think that our brain weighed the available evidence and then used logic to come to a conclusion. But what research has recently discovered is that we make an intuitive, emotional prediction about a scenario in an inaccessible, unconscious part of our brain and then our conscious cognitive processes in a totally different part of the brain try to explain if what we guessed was true. So the process is one of evaluating a variety of mechanisms that could result in a gut prediction rather than one of evaluating many possibilities to see which end result is the most likely.
This explains the movie results quite well. We want our gut instinct about how the movie will end to be correct, and we want our cognitive processes to enjoy watching the process to see how it happens. We predict what will happen next, and then see if we are right. We do that for the movie as a whole (the butler did it), and scene by scene. We want to be right about the ending, because that is a big prediction, but it's OK to be wrong about the scene by scene predictions as long as it still fits into the formula. Otherwise, we have no place in the brain to process the story and we get lost or bored.
This explains so many instances of confirmation bias. When we expect one brand to better than another, we focus on the attributes for which that brand is better and downplay the ones where it is worse. When doctors make a diagnosis, they look specifically for symptoms and conduct tests that will confirm their diagnosis rather than test it. When we instinctively like someone, we focus on the things they do that are positive and overlook their negative qualities. These same findings are found over and over again in context after context.
The fact that this is not optimal in the modern world, or not the way we would like cognition to work is really irrelevant. We have a lot of evidence that whether we like it or not, this is what happens. So why don't we start exploiting this knowledge to make better products, better systems, better user experiences?