Copyright: 2006
Publisher: MIT Press
ISBN: 0-262-03351-8

This "review" is more of a sketch of notes I took while reading this book.  Crandall, Klein and Hoffman deliver an intensely practical look into a realm of psychology that could be very difficult to comprehend.  The book surrounds the notion of Cognitive Task Analysis which is essentially a method of studying how people think.  Klein's work has surrounded primarily the way people make decisions (Sources of Power) however in Working Minds the authors also look at how people learn to do new jobs, how customers make decision purchases and a number of other ways that people think. 

 Resources

This book contains an incredible wealth of information and resources.  I would strongly recommend that if you are interested in learning such things as how people interact with computers (HCI), how to train people to use your software, how to transfer knowledge among your work force or any other field interested in how people think, get a copy of this book and use it as a jumping off point for your study. 

One web site mentioned in the book is:

http://www.ctaresource.com/

Data Collection Methods and Targets

Methods:

  1. Interview
  2. Observation
  3. Self-report
  4. Automated collection

Types of Targets (what to focus the methods above on):

  1. Time (events in past to how you might do it in the future)
  2. Difficulty (challenging for a novice? for an expert?)
  3. Realism (was it a real life event? is it likely to happen?)
  4. Generality (how specific is the scenario presented?)
Thoughts
  • The phenomenon being studied should overshadow the methodology.  Pick the method that will reveal the most information.
  • Differentiate between "end users" and "customers"
  • Concept Mapping - method of knowledge elicitation
  • Think out loud problem solving - method of finding out how experts vs. novices attack problems
  •  
Framing Questions

Some questions you should ask when beginning to undertake a CTA study.

  1. What issue or need do you plan to addres?
  2. What will you deliver?
  3. What sorts of people can tell you about the issue?
  4. What type of cognition do you need to know about?
  5. What type of situation can tell you the most about the issue?
 Features of Cognitive Landscape

Some things that you should consider when trying to determine how people think:

  1. Purpose or goals for a task
  2. How do we use our prior experience?
  3. Nature of the challenge (usual? physical?)
  4. Available tools to tackle task or solve problem (do  you have a calculator handy?)
  5. Team members (must consider cognitive collaboration if on a team)
  6. Organization (what is the organizational culture?)
Macrocognition

Thinking about thinking

Function

  • Naturalistic decision making
  • Sensemaking
  • Planning
  • Adaptation
  • Problem detection
  • Coordination

Processes

  • Maintaining common ground
  • Developing mental models
  • Mental stimulation and story building
  • Managing uncertainty and risk
  • Identifying leverage points
  • Managing attention
HCI (Human Computer Interaction)

GOMS (Goals, Operators, Methods, Selection) model.

David Kieras and Peter Polson developed cognitive simulations to study the cognitive work behind a task.

Envisioned World Problem - When you design a tool, it changes the way people do the task.  The design of the tool should take into account this fact.  What you are designing the tool for is a "moving target".

Scribner did some work involving "delivery route planning".

Information overload is caused by too much information being made available.  IT efforts should focus on high payoff areas of information for decision making purposes.

Information management is a means NOT an end!

Decision Centered Design

This method of design focuses on critical decisions primarily.  If you get the critical decisions right, you will likely get more of the others right... or at least it will be scalable. 

  • Identify the critical or difficult decisions to make
  • Identify WHY they are difficult
  • Determine how skiilled professionals make these decisions

Developers need to be careful not to become insensitive to the expertise of decision makers.

Customer Analysis

Do we know why our customers buy our products?  If not, why don't we study it?

What is our customers mental model when searching for a product in our category.  What problems are they trying to solve and what mental model are they using to search for a solution?

How do our customers know about our product?  Not just advertising but what is the relevant source, how reliable, how accurate, how up to date?

Study how decisions have been made in the past about related purchases.

Measuring new systems

Gains

  • Usefulness
  • Usability
  • Better immersion in the problem domain
  • Better recognition of situation
  • Ability to cope with unusual scenarios

Reductions

  • Gap between actual work and true work
  • Mental workload
  • Time and effort
  • Frustration level
  • Overconfidence / Mistrust of technology

Avoidance

  • Make work
  • Workarounds
  • Lack of understanding
Conclusion

Thinking about how you think can be frustrating, but it will force you to be more careful when you are trying to tell others how they should think about a problem.