Learning and performing new skills are easy; Problem solving is not

We have two minds

Not literally. That’d be creepy. Our brain is split into three primary functional areas:

The old (croc) brain separates all of our surroundings into three categories — consisting of edible, dangerous or sexy. It also regulates primal functions such as breathing, digestion and reflexive actions. The second part is the midbrain and it controls emotions. Finally, the third section is the new brain. The new brain is responsible for intentional, purposeful actions such as planning.

Our first mind, “system one”, controls human perception and behaviour. Meanwhile, the second mind is us and thinks its in control, even though it just chills in the back seat. System one works much faster than system two, only because it cheats by making approximations and doesn’t think things through. This is system two’s (the rational mind) role, but it can be lazy and just accept what system one is telling it. This is caused by the fact that system two “sleeps” until it is needed and system one’s judgements are usually good enough to get us by in a given situation. System two also kicks in when something that is “close enough” needs to be exact or when system one has conflicting responses.

Learning from experience is (usually) easy

We are very good at generalizing from experiences and observations and do so constantly and automatically. Learning from experience doesn’t require us to be aware of the fact we are learning. The more complex a situation is with a wide variety of variables, the harder it is for us to learn from it. Experiences from the people closest to us trump the opinions and experiences of others. Sometimes we may not learn the right lesson and don’t recall the actions we took to get ourselves there, creating an endless loop of failure until we clue in. Lastly, we we over generalize by making generalizations based on incomplete data.

Performing learned actions are easy

As we perform tasks or go somewhere over and over again, the process becomes either fully automatic or semi-automatic. “Automatic” is another term for routine. Routine tasks require little to no conscious attention, meaning we can multitask when the task we are performing is simple.

Performing novel actions is hard

The first time we learn a complex task, our rational mind is fully engaged, putting us under a lot of mental strain which leads us to become very overwhelmed, as in learning how to drive. Most daily tasks though involve a mix of automatic and semi automatic components. When designing an experience, whether it be software or the web, features should target system one’s automatic response or have the ability to quickly become automatic.

Problem solving and calculation is hard

Because of our new brain (cerebral cortex), we have the ability to solve problems and not just rely on instinct. We can think our way out of situations and come up with creative solutions on the fly. Learned experiences are easy because they require far less conscious attention where solving problems trains our short term memory and forces our brain to compete for scarce attention resources. Our rational brain has a very limited storage capacity and is much slower than our old brain. System one is like a 64-bit processor while system two is like a Pentium 3.

We aren’t good at calculations and our short term memory isn’t good, so performing math entirely in our heads is nearly impossible unless part of the calculation has become automatic, such as memorizing the multiplication tables. When solving problems, people frequently use external aids such as writing down results or drawing diagrams.

Implications for user interface design

Sometimes its necessary to push oneself to strain their brain, but they don’t appreciate it when a problem they didn’t expect is thrown in front of them. People are driven by goals and are using computers to meet those goals and must put all their attention on them.

Some design guidelines

· Indicate system status and users’ progress toward their goal.

· Guide users toward their goal. Provide a clear information scent that a user can follow to reach their goal. This will prevent a user running out of short term memory capacity.

· Tell users explicitly and exactly what they need to know. Don’t expect them to deduce information that you may think that they know.

· Minimize the number and complexity of settings. Not all people are good at customizing a lot of settings.

· Don’t make users diagnose system problems.

· Make the system familiar. Use concepts, graphics and terminology that your target audience would use and recognize.

· Let people use perception rather than calculation. Users will have a higher success rate if they don’t have to do complex calculations to complete a task.

· Let the computer do the math. That’s why it was invented.