|
|
 |
|
Thinking About Thinking
© Esther Derby 2000-2002
This article originally appeared in STQE Jan/Feb 2001.
As development managers, test managers, and testers, we think about and make decisions every day. We face all sorts of complex problems: ferreting out the fault behind a failure, selecting a debugging tool or defect tracking system, choosing the best candidate to fill an open position, analyzing potential problems that will put a project at risk. Some of these problems call on specific technical skills and knowledge. All of these tasks depend on our cognitive abilities, our ability to think. While we can take classes to increase our technical skills, how do we go about improving our
thinking skills?
I'd recommend two books-one recent, and one that's been around for a while-as a good place to start. In
The Logic of Failure, author Dietrich Dörner reports on the cognitive and psychological tendencies of people solving complex problems. Dörner, a professor of psychology at the University of Bamberg, focuses on "system" problems, in which the problem solver faces many inter-dependent variables and incomplete information. His book will give you insights on common cognitive mistakes-traps that you can recognize and avoid.
The more recent of the two books, William J. Altier's The Thinking Manager's
Toolbox, further illuminates the landscape of thinking by providing straightforward, logical processes that are useful for decision making and certain types of problem solving. Altier focuses on more clear-cut situations, and on driving for clarity in ambiguous situations.
Both authors share the same goal: helping you be a better problem solver. They stress the importance of recognizing the situation you're in, choosing an appropriate problem-solving strategy, and having the right thinking tools.
Being Merely Human
Okay, here's the big news: We're human and we have some cognitive and psychological limitations. As humans, most of us just don't do well solving highly complex problems under pressure. Where do we go wrong?
The Logic of Failure offers some answers. I first read this book a couple of years ago...and it's still just as interesting the second and third time around. The author describes a series of experiments that involved having normal, intelligent people (for the sake of argument, let's say that's you and me) solve complex problems in computer simulations.
As you might guess from the title, the experimental subjects' results were less than stellar. They showed characteristic tendencies that always lead to poor results:
- Failure to recognize networks of interdependent variables, where a change in one variable will affect many other variables
- Consistently underestimating exponential growth (People are often surprised, for example, that taking 1, doubling it, doubling it again, and so forth leads to more than 4 billion in only thirty-two steps
- Failure to formulate clear goals
- Failure to consider ripple effects and long-term consequences
- Retreat into familiar, solvable problems, even when they are not the problems that need to be solved
- Failure to reflect on the outcome of decisions and actions
Do any of these sound familiar to you? I recognize some of these behaviors in managers and software professionals I've worked with over the years. I'm reminded particularly of one vice president of software development who insisted on
personally approving each and every reimbursement request over $5.00 in an organization of a thousand people-while multi-million dollar projects were spinning wildly out of control. Reading Dörner's book doesn't make that kind of behavior less amazing, but it does make it more understandable. This VP wasn't stupid, merely human: faced with an overwhelming mess, he retreated to solving tractable
problems, even if they were unimportant problems.
The Logic of Failure paints a pretty dismal picture of our human abilities to solve problems in complex systems. Software systems, software projects, and software organizations are pretty complex systems; should we just pack up and go home? Or resign ourselves to muddling through?
Well, consider that there were some subjects who did do well-and they showed common patterns in how they approached problems. They tended to ask more questions. They were less likely to reverse their decisions in favor of an opposite course of action, but also less likely to hold onto decisions when circumstances dictated a change. They reflected on the results of their decisions and their actions. They showed an ability to recognize which problems were important, and choose appropriate strategies to deal with them.
This isn't the stuff of genius, says Dörner, but of learning through experience and learning through reflection. "What matters is not, I think, the development of exotic mental capabilities," offers the author. The one thing that does matter, he argues, is the development of our common sense. According to Dörner, "Thinking about our own thinking...can make us better problem solvers."
Developing a Thinking Discipline
The Thinking Manager's Toolbox by William J. Altier is a how-to guide for effective problem solving and decision making. This 1999 book is explicit about analytical thinking processes. While it doesn't cover every human cognitive shortcoming, it provides clear guidance for setting goals and making tradeoffs, choosing among alternatives, identifying potential future problems (a.k.a. "risks"), and tracking down the causes of problems you already have.
Altier devotes the first part of the book to "fundamental" analytical processes, with detailed descriptions, examples and tips for:
- Situation Analysis, for understanding what kind of thinking is needed
- Decision Analysis, for making various types of choices
- Potential Problem/Opportunity Analysis, for identifying potential detrimental and beneficial conditions
- Problem Analysis, for finding the root cause of a problem, and
- Implementation Planning, which is a special application of decision analysis
The second half of the book, "The Advanced Toolbox," talks about how to adapt the fundamental processes to your own situations. It also covers using Scenario Planning, which addresses one of the weaknesses Dörner identified in his book-the tendency to deal only with immediate problems and ignore possible future problems, including those caused by the "solution" to yesterday's problem. The appendix provides worksheets and templates to use in all the processes described.
Why do we need to follow a defined, step-by-step process to think? Because an explicit process, says Altier, "enables people to take mental steps together." This is one of the author's key ideas: in the work of managing software development and software testing, we can't just get by on our own mental juice. We need to bring our teams, our customers, and our peers into agreement on goals, decisions, and problem causes. And we need to engage others in planning, identifying and mitigating risk.
How do the thinking processes described in Toolbox apply to the wonderful world of software?
Ever had to negotiate a compromise between delivery date and functionality for a software product? Ever had to choose an off-the-shelf accounting system or bug tracking system? Planned the implementation of a release? These are all thinking activities that are common in software and well suited to the analytical approaches in Altier's book.
Expanding Your Toolbox
In keeping with Altier's concept of a toolbox, look at analytical techniques like those in his book as just one tool in your collection. A skilled problem solver will also develop the skills that Dörner's successful subjects exhibited: the ability to understand the internal dynamics of a system and its tendencies over time, and build habits of reflection. A skilled problem solver will assess the situation and choose an appropriate approach.
Sound like a challenge? Well, yes, it is. It requires effort to reflect our own thinking, to disrupt our habitual patterns of thought, follow all the steps of a logical process, or wrap our brains around networks of causation and vectors of exponential growth. And it's not always comfortable to look back and see where our thinking was inadequate and subsequent actions were way, way off.
Dörner and Altier suggest we apply more rigor in the way we use our brains-that we think about
how we think. And if that allows us to help short-circuit the logic of failure, then these two books seem like a good investment to me.
The Logic of Failure: Recognizing and Avoiding Error in Complex Situations
By Dietrich Dörner
Published by Perseus Press 1997
The Thinking Manager's Toolbox: Effective Processes for Problem Solving and Decision Making
By William J. Altier
Published by Oxford University Press 1999
|
