In an earlier blog I spoke of a tongue-in-cheek desire to assemble a catalog of perceptual patterns humans are likely to get wrong. The example discussed was a visual perception issue—the trajectory of a bouncing ball. I was imagining the catalog as a humorous compendium of cognitive pattern recognition human shortcomings—a handy fallibility reference. However, as seemingly trivial as the trajectory of a bouncing might be, I was nevertheless intrigued by how we can use visual pattern recognition to expose or render visible the social patterns of human behavior.
I think that one of the many strengths of SenseMaker is its ability to visually display these socially derived patterns. As I was thinking about how I can more effectively use the visual aspects of the software to the benefit of myself and clients, I came across an interesting article about how we perceive the Necker Cube. For those that aren’t familiar with the Necker Cube—it’s one of many stable images that can be seen, or perceived, in two different ways (see the link below to take the Necker Cube visual test).
The experiments conducted by Jürgen Kornmeier, Christine Maira Hein, and Michael Bach and reported in their 2008 article, Multistable perception: When bottom-up and top-down coincide in Brain and Cognition yielded a variety of interesting results. I was intrigued by the finding that if the Necker Cube image was flashing it was easier, and a whole lot faster, for human subjects to see both the perceptual alternatives. There was also some evidence that subjects skilled in meditation (Buddhist Monks in this case) are able to stabilize one of the image alternatives for extended periods.
In this experiment, flashing the image made an alternative perception easier for the test subjects to access. I wonder if the stimulation of image flashing would be useful in more effectively exposing pattern differences in visual data sets. It seems to me that anything we can do to both enhance our pattern recognition capacity and expose patterns which are invisible because we don’t know how to see them would be worthwhile.
My status as a guest blogger is coming to a quick end, but the last couple of weeks have been fun. Every time I sat down in front of the computer to write a blog, Dave’s admonition came to mind—the one about knowing more than we can say and saying more than we can write. Ciao
To take the Necker Cube test and read more
http://scienceblogs.com/cognitivedaily/2008/11/seeing_illusions_in_two_differ.php#c1218600
Comments (1)
It was great to connect with you (asymmetrically) Craig by following your blog and we are overdue for a phone call. Perhaps this week upcoming.
Your reference in this blog post to the Necker Cube and commentary about it made me think about one way that I use SenseMaker when building client reports. One of the things I do is to include screenshots of the patterns I see and then annotate them in my reports (common for most of us working with SenseMaker as you saw in the recent project we worked on together). What I often do is carefully specify the same size and crop of certain images and then show different perspectives of those same images based on the interpretation given to the set of stories. For example I may take the same pattern view for new hires, junior employees, mid-term employees, and long-term service employees. Then if I place these identical images (say for example on attitudes towards leadership, level of engagement, etc.) in the same positions on a series of PowerPoint or Keynote slides I can then sequence the slides very quickly and the only thing that changes is the pattern itself and associated statistical indicators that are selected for the view (ie. mean, median, percentiles). In the sequencing I see subtle and not so subtle shifts and then look at the static view to make sense of the shift I saw during the rapid slide sequencing. Accessing the stories behind the pattern differences then often reveals thematic differences that help understand the more nuanced aspects of the culture or work environment. Perhaps not an equivalent example of the Necker Cube test however the flashing seems to play a similar role in enhancing pattern detection.
PS I had some fun with the test in trying to train myself to see the alternate perspective with each flash. Near the end I was focusing on one line trying to alternately see it in front then behind and it seemed to work for a while but difficult to hold for a long time.
Posted by Michael Cheveldave | January 25, 2009 6:07 AM
Posted on January 25, 2009 06:07