Behavior design patterns

In my lecture yesterday, I presented on 16 various forms of influence
strategies. One thing that Jeff Heer and I were discussing was the
idea of identifying behavior design patterns from looking at existing
systems. Which strategies complement one another? Which are used for
what kind of situations? This would be similar to the programming
design patterns and would require a significant set of existing
applications with tags of each strategy in use and then distilling
patterns from those.

Guest lecture at Stanford's interaction design studio class

Click here to download:
BehaviorDesign-cs247-20110208.pdf (1.07 MB)
(download)

Today I gave a guest lecture in Jeff Heer and Sep Kamvar's interaction
design studio class, cs247
. The topic was behavior design. In it, we
discuss the two sides to behavior design: (1) the psychology of
influence and motivation and (2) designing behavior into the user
interaction. The slides are below. I used a lot of BJ Fogg's work as
well as influence strategies from Cialdini and several practical
examples.

This week's reading

Click here to download:
Andrew_PT07.pdf (114 KB)
(download)

This week's reading for our class on Designing Technologies for Casual Learning is embedded below.  The paper feels cutting-edge when it comes to the actual suggestion tactics but dated when it comes to the review and 3 technologies used.  It shows how much such tools have proliferated.

Andrew, et al, use the following suggestion tactics summarized by Fogg:

  • Reduction: Making a complex task simpler
  • Tunneling: Guided persuasion; giving control over to an expert
  • Tailoring Customization: providing more relevant information to individuals
  • Suggestion: Intervene at the right time with a compelling suggestion
  • Self-monitoring: Automatically tracking desired behavior
  • Surveillance: Observing one’s behavior publicly
  • Conditioning: Reinforcing target behavior

Augmented Relaxation

I've been thinking for the past few months about a trend towards persuasive technologies that try to calm us down, reduce anxiety, and reduce stress.  I've been referring to this emerging sub-field as "augmented relaxation".  This includes apps that force you to take a break while using your computer, guiding meditation, remind us to attend yoga, or guide contemplative practices such as meditation.

I have been working on respiration in particular.  "Persuasive respiration" is the phrase I've been using (but I don't love it) to describe how technology can help augment our respiratory habits.  Linda Stone has referred to the problem of 'email apnea' that occurs while users use their computers.  This is one instance of the problem my research aims to eliminate.

There is a lot of motivation to address the problem of the H's state in 'HCI' (human-computer interaction).  Affective computing normally uses the user's affect to change the UI.  In our case, the technology is interested in ensuring the user's affect is 'optimal'.  Not 'happy' or 'effective' but, rather, calm/clear.  Respiration is a key lever of the autonomic that controls our stress level and, indirectly, clarity of thought and, I would speculate, creativity/long-term productivity, etc.

My dissertation aims to develop the underlying principles that help ensure their users are cool, calm, and collected.  I argue that this improvement in turn improves productivity, creativity, innovation, and long-term effectiveness.

Metrics for self-shaping

As a cyclist, there are several ways in which individual's track their activity.  This variety of metrics is present in other domains as well, but cycling provides a particularly clear array of possibilities.  The first possibility is simply distance.  How many miles do you ride at once, in a week, a month, or a year?  This is an often used but rather superficial metric.  More telling is the combination of a distance metric and a rate metric, average speed.  Covering 10 miles in 30 minutes is much different than in an hour.  

Two other external metrics exist and are increasingly common in cycling training.  Power meters, measuring the amount of energy being put into pedaling, are popular.  In addition, cadence, the number of pedal rotations per minute, is another metric, one which is often found on exercise bikes.  Lance Armstrong, for instance, is famous for maintaining a cadence of 100 pedal rotations per minute, 20 higher than many of his competitors.  

Personal metrics also exist, these can be subjective and objective.  Objectively, measures like heart rate, maximum oxygen usage, and muscle acidity, are important, though not always easily obtained.  Subjective exertion is a last common metric, and, for experts, is often the best. Lastly, subjective evaluations of cycling, such as its enjoyability, challenge, meaningfulness, and utility are also potentially relevant to goals, even cycling goals.

What metrics do you use to pursue your self-shaping goals?  Are others available?  Do these metrics support internalization (moving from objective metrics to a reliable subjective metric) or low-maintenance self-monitoring, so that you can seamlessly monitor progress?

 Matching metrics to goals is also important.  If my goal is to maintain a level of aerobic fitness, then I should use my metrics differently (maybe even use different metrics), than if I am seeking to maximize my performance in a sprint.  .

Is learning just a behavior we want to create in users?

This week, as part of the course I'm teaching with Dr. Goldman called
The Design of Technologies for Casual Learning, we are incorporating a
significant persuasive element to it. These are the two readings for
the week. I'll be including in the lecture notes on Cialdini's
'Influence' and Fogg's Behavior Grid.

What's interesting about this week is that we are designing learning
opportunities endogenous to the gaming experience and also using
goal-setting to create behavior change. Is learning a behavior? What
is the impact on learning technologies if we treat learning as a
behavior we want to trigger?

Click here to download:
a40-fogg.pdf (521 KB)
(download)

Click here to download:
Persuasive09-consolvoEtAl.pdf (196 KB)
(download)

Couch-to-5k: A persuasive exercise iPhone app

http://learncasually2011.posterous.com/couch-to-5k
This app is different from many related iPhone apps, having a clear
goal in mind and targeting a specific user. The app clearly leverages
the "commitment" influence strategy. Take a look at the screenshots.
It not only reminds you to cool-down but does so at the right time.

The peer influence is not as strong: it's optional and is more like a
"show off" thing that some users might use but isn't central to the
app.