Macrocognition

Welcome back! If you celebrate the holidays, I hope you all had a good one. We’re going to be starting off with a shorter article to help ease in to the new year, but it’s no less interest or useful.

I also want to point out that there is a new sponsor. Adaptive Capacity Labs is made of folks you likely already know, some of which I’ve featured here. I’m happy to introduce them as the new sponsor!

Macrocognition

This is an article by Gary Klein, Karol Ross, Brian Moon, Devorah Klein, Robert Hoffman, and Erik Hollnagel from IEEE Intelligent Systems. They go over why there is value in studying macrocognition as well as an overview of how it differs from other ways of studying and approaching cognition. What I find most useful about the article is that they provide enough information to use or at least begin to use the macro cognitive view as a lens through which we can look to evaluate our systems and processes.

Macrocognition is contrasted with microcognition. The big differences are that one happens in real life (“naturalistic” environments) whereas the other happens in the lab. In the lab environments, typically researchers try to isolate a small part of the whole to better understand it.

Macrocognition is not necessarily a “better” approach to research, but it is one that has received less attention and one that I think can be more useful to apply as designers and maintainers of software systems.

The authors point out that if we create complex cognitive systems (and of course as software systems people we do)m that if we do so with a flawed idea of how cognition works then those flaws will also show up in the systems we design. This means that the system may hurt instead of help.

Macrocognition involves fuzzy goals, which is a lot closer to how we typically work, as opposed to the small, very well defined goals of microcognition. Further, the authors explain that in addition to happening in the real world,

“macrocognitive functions are generally performed in collaboration… and usually in conjunction with computational artifacts.”

This means we can use the macrocognitive view to look at how a team might function, instead of just an individual, for example, as we explored in issue 40.
The authors provide us with a list of macrocognitive functions and supporting processes that apply to individuals, teams, and technologies. These include:

• Naturalistic decision-making
• Sensemaking
• Planing
• Adaptation
• Problem detection
• Coordination

With supporting processes like:

• Maintaining common ground (see issue 25 for more on common ground)
• Developing mental models (see issue 61 for more about the strengths and problems with mental models)
• Uncertainty management
• Attention mangement

The authors present those lists in a somewhat circular diagram since they all can happen at the same time or many of them need to be accomplished at once. I list them here linearly in no specific order. Additionally, the authors point out, any one of these could be a goal itself or just a means to an end. For example, problem detection might be a desired goal for medical care or might help replanting.

I think that just knowing a contrast to the oft discussed lab environment view exists, along with some of these functions can help broaden our perspective as we develop processes and systems. Instead of asking questions like “how does this help with memory?” or “how does this help vigilance?” (a subject I’ll be addressing in a future issue), we can begin to ask questions like “will this work for teams solving problems as well as individuals?” or “will this help individuals make decisions or replan?”

Generalizing from studies that involve firefighting, critical care nursing, and military decision-making, they found that people make decisions using recognition strategies and very few by just comparing options in an analytical way. In addition people typically are evaluating these options by simulating them in their head not just sort of robotically comparing features. Further, as people became more experienced they spent less time contrasting options and more time examining the situation.

These are all macrocognitive function that our tools and systems we design can help support.

Takeaways

• Creating cognitive systems with a flawed view of how cognition works may cause us to create systems that hurt rather than help.
• Understanding how cognition works in the real world can help us create better and more useful systems and tools that support the humans using them.
• Macrocognitive functions include things like: planning, coordination, and adaptation.
• Macrocognition is looking at cognitive function in the real world and not a lab.
• Many macrocognitive functions can happen at once and many may be needed to accomplish a particular goal or they can be accomplished as an end themselves.
  • This means there isn’t really a linear relationship between the various functions. Unlike some views of microcognition, we can’t really see these as steps that we go through.
• As individuals gain more experience in their domain, they spend more time on examining the situation and less time examining options.