#38 from R&D Innovator Volume 2, Number 5          May 1993

Tactics for Creative Problem Solving
by Steven H. Kim, Ph.D.

Dr. Kim, currently visiting at the University of Virginia, has served on the faculty of the Department of Mechanical Engineering and as director of the Knowledge Systems Program at the Massachusetts Institute of Technology.  In 1989, he received the National Science Foundation's Presidential Young Investigator Award.  He is author of Essence of Creativity: A Guide to Tackling Difficult Problems, Oxford University Press, 1990.

For a researcher, creative thinking is essential.  Fortunately, we can take deliberate steps to enhance creative problem solving.  In the preparatory phase of a difficult task, we must carefully consider the underlying problem and have a clear understanding of the goals. 

Too often we tackle tasks without an adequate definition of the true problem, and discover too late that we have solved a trivial problem.  This article takes a very general view of a problem as any goal-oriented task requiring decisions for its fulfillment. 

Once the problem is properly defined, I find it’s useful to partition the "idea-enhancement" practices into three phases:  input, processing, and output.

The Input Phase

Quality thinking requires quality input, or information.  You must expose yourself to a rich spectrum of stimulating ideas from colleagues, journals, and other sources.  Since it's usually not clear where to seek inspiration to solve a difficult problem, so one recourse is to skirt the perceived boundaries of the main problem and delve into related fields.  The history of science demonstrates the fruitfulness of such an interdisciplinary approach.

The inspiration to solve a problem doesn't necessarily arise from technical know-how.  Studies of innovation in the industrial sector indicate that new products often originate from potential users of a product.  For example, Chester Carlson, a patent attorney, invented a way to copy documents to eliminate the errors and inconvenience of retyping patent applications.  This led to xerography—and the rest, as they say, is history. 

The Processing Phase

During the processing phase, new ideas are generated.  This phase requires a judicious mix of "foreground" and "background" mental processing.  In the former period, we use periods of solitude for quiet contemplation.  In the latter period , ideas arrive "in a flash," while we are driving or drifting off to sleep, whenever the mind is not consciously engaged in a difficult foreground task.  These “subconscious” solutions to our toughest problems are often unheralded, but we are unlikely to encounter great insights while playing a strenuous game of tennis or struggling with insomnia.

A number of techniques have been developed to enhance the creative process.  Among the most widely known are morphology, brainstorming, and synectics.

Morphology systematically identifies all combinations of relevant options:

•  Identify the problem domain.

•  List all relevant attributes or dimensions.

•  Identify all possible combinations.

•  Evaluate each combination.

For instance, to determine the best condition for extracting a pharmaceutical from solution, we consider five solvents, under three temperature and two pH conditions.  Thirty combinations must be examined to find the best one. 

This method requires an enumeration and consideration of all potential solutions.  Unfortunately, some modern challenges have so many possible combinations that computational constraints usually preclude a complete search.  The morphological approach is therefore appropriate only for problems of limited scope, where a manageable number of variables can be identified in a relatively straightforward way.

Brainstorming is the well-known spontaneous generation of ideas by a group of people, some of whom may know little about the problem.  The object is to use free association to generate all possible solutions.  Each proposal, however outlandish, is recorded during the session and not evaluated until later.  However, this technique has limitations.  While the quantity of new ideas may be high, there is no pressure at the generative stage to limit ideas to those with high quality (those that are economically or technically feasible, for example).

Synectics  is another small-group method that can be used by individuals with diverse backgrounds.  Similar to brainstorming, synectics's structured approach uses these principles:

•  Make the strange familiar.  Define the problem so everyone understands it.

•  Make the familiar strange.  Use analogy to view familiar objects in a new light. 

For instance, a person may imagine herself as the product (e.g. a new apparatus) of the research program.  She would imagine actually being that apparatus, in operation.

The synectic approach identifies the fundamental concepts of a situation; it does not emphasize obvious characteristics.  This technique is especially useful for producing radical departures from traditional paradigms. 

Output Phase

During the output phase, we record ideas—whether they are still half-baked or complete.  Sometimes the complete solution appears suddenly from an unknown quarter, as we merely try to clarify it on paper, or try to explain it to someone.  At other times, new ideas occur while we are writing down a partial solution or preliminary results.  Perhaps, while reading a preliminary draft, we find deficiencies and opportunities that pave the way to the next draft, with a better statement of the solution.

By whatever means, the solution must be expressed so others can appreciate it.  A marketing expert, for instance, will not understand your idea if you explain it in chemists' jargon.  Thus a good idea may be ignored just because you could not express it suitably.  If you state your idea in technical terms not appreciated by the reader or listener, you haven't really expressed it.  Who needs to know about your idea?  How do you get that individual to buy into it and pursue it?  You may have to play the game of "language-politics" to package your idea in a way palatable to your customers and other stakeholders.

Some aspects of creative problem solving will enhance all three of the above phases.  If your colleagues are supportive and interactive, they will supply a critical mass of experiences and attitudes to argue through a strategy or provide information relevant to the project. 

Often when one person suggests an idea, it is refined by second, then further developed by someone else.  As problems gain complexity, problem solvers should rely to a greater extent on such synergies among colleagues.

Directed Refinement

For especially difficult problems, a more focused approach is needed to handle the myriad alternative avenues of investigation.  The procedure I call “directed refinement” has these stages:

Breadth-first search.  Globally scan the problem and develop general tactics to approach it.

Depth-first search of the identified path.  Follow the narrowly defined path until specific information arises to indicate a change in direction.

Evaluation.  Evaluate the results to determine progress towards the goal.

Decision.  Decide if the goal requires a fundamentally different approach.  If so, determine whether a breadth-first or depth-first search is again necessary.

By using directed refinement, the problem solver can keep track of unexpected developments and continually refine strategy to expedite the solution.

The nature of a complex problem may change over time.  While working on a problem, you continually gain new concepts, knowledge, and techniques.  Which ones will prove most useful in the future?  If you have a choice, pursue general or enduring topics.  For instance, it may be more useful in the long run to understand the concept behind a method than to be technically super-competent with that method.

Some researchers intuitively follow the types of procedures I have just described.  But many people will benefit by considering these tactics—and diligently applying them to the toughest problems.