A rule of thumb approach which may be more prone to error but faster in problem solving is a

    Problem Solving & Creativity

s
A rule of thumb approach which may be more prone to error but faster in problem solving is a
Back to Cognitive Psych Home page 
A rule of thumb approach which may be more prone to error but faster in problem solving is a
Back to Topics 

Problem Solving
  • Overview
  • Problem Typologies & Internal Representation
  • Problem Solving Approaches
  • Factors that Influence Problem Solving
Creativity
  • Definition
  • Approaches to the Study of Creativity
Problem Solving

OVERVIEW

Chi and Glaser (1985) define problem solving as a situation in which a person is trying to reach some goal and must find a means for arriving at it. Interest in the study of problem solving originated in the work of the Gestalt Psychologists in the early part of the 20th century. As you might imagine, they saw problem solving in perceptual terms -- as a novel rearrangement of the perceptual elements of the gestalt. Thus problem solving was seen as a discontinuous process in which one achieves "insight" into the solution by the sudden rearrangement of elements. This view is supported by the many times you have had an "ah hah" experience in suddenly solving a problem that you have been struggling with -- perhaps for days or longer. Below is an example of just such a problem. Used by Silvera (1971) in establishing the incubation effect, it is known as the cheap necklace problem. Try and solve it.
    You are given four spearate pieces of chain that are each three links in length. It costs two cents to open a link and three cents to close a link. All links are closed at the beginning of the problem. Your goal is to join all twelve links of a chain into a single circle at a cost of no more than fifteen cents.
A rule of thumb approach which may be more prone to error but faster in problem solving is a

Don't be discouraged if it took you a while. Only 55% of the subjects were able to solve the problem within the first half-hour. Some subjects worked on the problem briefly and then had breaks of either a half-hour or four hours. Subjects with breaks solved the problem more quickly, with the group having the four hour break doing best. The outcome supports the phenomenon of incubation for those problems that lend themselves to insightful solutions. It turns out that not all problem solving is discontinuous in nature. Some types of problems require incremental (continuous) solutions. Performance improves with increasing attempts or in steps.

The study of problem solving waned during the behaviorist period. Behaviorism, as you know, saw problem solving in terms of trial and error with reinforcement strengthening the correct response in an incremental way. The emergence of Cognitive psychology with its emphasis on the information processing approach, has produced renewed interest in problem solving as a cognitive phenomenon. For the most part, the information processing approach has adopted an incremental view of problem solving -- problems are solved in a series of steps.

Irrespective of whether one adopts a discontinuous or incremental model of problem solving, all agree that the most critical factor in understanding the problem is constructing an adequate internal representation of the problem. Since a first step in constructing an adequate representation might well be characterizing the kind of problem one faces, we look first at typologies of problems suggested by Greeno (1978).

Return to top

PROBLEM TYPOLOGIES AND INTERNAL REPRESENTATION

Greeno (1978) has suggested three basic forms of problems each associated with a particular cognitive operation or skill necessary to solve problems of that type.
  1. Problems of Inducing Structure --

  2. determining the relationship among several given elements of the problem. Example: analogy problems
    DOCTOR:PATIENT::LAWYER:_________

    Requires three kinds of skills:

    • Attribute discovery or encoding
    • Comparison of encoded attributes
    • Evaluation
  3. Problems of Transformation --

  4. finding a sequence of operations that transform the initial situation into a goal state. Typically, the permited operations are limited and constraining. These kinds of problems lend themselves to formalization (expressed in symbols with rules) Examples:
    The Five-Handed Monster Problem (click to view)
    A rule of thumb approach which may be more prone to error but faster in problem solving is a

    The Hobbits-and-Orcs Problem (click to view)
    A rule of thumb approach which may be more prone to error but faster in problem solving is a

    Skill required: Means-End Analysis

    • Determine difference between present state and goal state
    • Select a legal operation that moves you closer to the goal state
    • Evaluate Progress
  5. Problems of arrangement --

  6. problem is to arrange some element according to some criterion. Example: The anagram problem

    Arrange the letters AIFMA to make a word

    Skill required: Constructive Search

    • Systematically examine reasonable combinations until solution is found
    • Use above to generate partial solutions
Further Requirement for Understanding a Problem Greeno (1977, 1991) suggested three requirements for understanding a problem:
    1. A coherent representation

    2. --all the parts are connected in a way that makes sense
    3. Close correspondence between internal representation and elements of the Problem

    4. --the representation "maps well" onto the elements in the outside world
    5. Elements of the problem must be related to background knowledge

    6. --links to existing knowledge provide the source for solutions
Specific Methods of Representing the Problem

Matlin (1998) suggests a number of methods of representing the problem:

    1. Symbolic Representation (Algebraic, Boolean Algebraic)
    2. Lists
    3. Matrices (cross classified lists)
    4. Hierarchical Tree Diagrams
    5. Graphical Representations
    6. Visual Images
Return to top

PROBLEM-SOLVING APPROACHES

Two very commonly-used strategies in solving problems are the algorithmic and heuristic.
    Algorithm A finite series of steps which if faithfully performed will always result in a task or process being completed -- in this case, the problem will be solved. In problem solving, this approach is sometimes called a "brute-force" solution because all possible rearrangments are tried. Heuristic A rule of thumb -- selective searches involving looking at only those portions of the problem space that are most likely to produce a solution.
We can see an example of each approach in considering the solution to a classic problem in problem-solving, computer science, and artificial intelligence -- the cryptarithmetic problem: The Problem: In the example below, each letter represents a single numeral and the correspondence between the numerals and letters is one-to-one. Your task is to deduce the correspondence such that when numerals are substituted for letters, the resulting addition problem is mathematically correct
   DONALD
 +GERALD
  ROBERT
The algorithmic solution is to simpy make all possible assignments of digits to letters until the correct solution is found. Notice while such an exhaustive search might be quite difficult for humans, it would present no problem to a computer.

The heuristic solution, on the other hand, starts with what we know about numbers. Supposing I tell you the D=5 as a hint. You now deduce that T=0, that a carry of 1 is generated, therefore R has to be odd since the same number added to itself produces an even sum -- so L+L+1 has to be odd, etc. One very important heuristic to emerge from problem solving studies is the Means-Ends Heuristic.

The Means-Ends Heuristic

Characterized by two components:

  1. The overall problems is divided into a number of smaller subproblems with a subgoal for each of the subproblems,
  2. You seek to reduce the difference between the initial state and subgoal state for each of the subproblems.
The means-ends heuristic is quite compelling and sometimes is counterproductive in problem solving. There are occasions when it is necessary to implement a step that temporarily leads away from the goal. The successful solution to the Hobbits-and-Orcs problem required steps where you actually returned some creatures from the far bank to the near bank, thus moving away from the goal of all creatures on the far bank. Usually, overcoming the means-end-heuristic leads to fairly rapid solution of this problem.

Newell and Simon and General Problem Solver

Allen Newell and Herbert Simon (Newell & Simon, 1972; Simon, 1995) developed a theory of problem solving that conceives of:

    • The Task Environment --

    • the true representation of a problem in an objective and neutral way
    • The Problem Space --

    • the internal representation of the problem -- usually incomplete
The problem space is conceived of as resembling a semantic network. Each node represents a particular state of knowledge. The nodes are linked by cognitive processes called operators that allow movement from one node to another. So problem solving consists of moving through nodes of the problem space.

In an attempt to support their theory, Newell and Simon developed a computer program that simulated human problem solving. These researchers had human subjects think out loud as they solved problems (among them, the cryptarithmetic problem) from which they created written protocols. Processes and strategies that emerged from the protocols were encoded into the program called General Problem Solver or GPA. The program, intended to solve a wide range of problems incorporated the human strategy of working with subproblems, of seeking steps that would reduce the distance to the goal, of applying the appropriate operator to bring this change about.

The program was most succussful at solving the kinds of problems that lend themselves to formal representation (like Hobbits-and-Orcs, Five-Handed Monsters, Towers of Hanoi). The point to computer models of problem solving isn't so much to turn problem solving over to computers as it is to force theorists to clarify and increase the precision of their theories. Work in this area continues.

Return to top

FACTORS THAT INFLUENCE PROBLEM SOLVING

    Expertise --
    Since problem solving depends do critically on a well-developed internal representaton of the problem, it is not surprising that expertise, with the many additional links it provides, is an important factor. Among the specific advantages, one would have to list:
    • Increased knowledge base
    • Superior memory for facts, event, and processes in the area of expertise
    • Accuracy of representation
    • Enhanced problem-solving approaches
    • Ability to elaborate on initial states
    • Speed and accuracy
    • Metacognitive skills
    Mental Set --
    An inability to adopt new strategies in solving problems operates against finding solutions.

    Functional Fixedness --
    An inability to assign new functions and roles to elements of a problem does likewise

Return to top Creativity

DEFINITION

Creativity as a concept is very difficult to define. Like deja vu, it is impossible to manipulate and produce on demand in the laboratory. Like beauty, its existence may lie in the eye of the beholder. However, like pornography, while difficult to define, we know it when we see it. Clearly, however, it is part of problem solving. Among the qualities that characterize creativity are:
  • The solution has novelty and usefulness, either for the individual or society
  • The solution requires adopting new ways of thinking about the elements of the problem
  • The solution frequently results from intense motivation or persistence
  • The solution comes from clarifying a problem that was initially vague
APPROACHES TO THE STUDY OF CREATIVITY

Given the difficulty in defining the concept of creativity, it's not surprising that the methods of studying it are quite diverse and may not strike you as having, in all cases, face validity. Among the approaches are:

  1. Guilford's Divergent Production Test

  2. In this test, subjects are encouraged to make as many and as varied responses to test item as possible. The greater the number, the more creative the person is assumed to be. The two items below suggest how the test works:
    • Many words begin with an L and end with an N. In one minute, liist as many words as possible that have the form L_______N.
    • Suppose that people reached their final height at the age of 2, and so normal adult height is less than 3 feet. In one minute, list as many consequences as possible that would result from this change.
  3. The Remote Associates Test

  4. This test was devised Mednick and Mednick (1967). In the test, each item consists of three words which must be linked together by a single word. Some sample items:
    • cookies sixteen heart ______: The correct associate is "sweet."
    • poke go molasses ______: The correct associate is "slow."
    • surprise line birthday ______: The correct associate is "party."
  5. Consensual Assessment Technique

  6. A method suggested by Teresa Amabile who suggests that creativity is a property of products rather than people. In this approach, a product is considered creative if a group of expert obervers judge it to be so.
Correlations for divergent production scores as well as RAT scores with other measures of creativity are only moderate, so it's not quite clear what's being measured. Correlations for the Consensual Assessment Technique are high but one would expect that given that creativity is being operationally defined, i.e., creativity is what a group of experts say it is.

One final word: The role of incubation in creative problem solving should not be underestimated. In both the Gestalt and contemporary literature, one finds support for the notion that time spent away from a problem appears to enhance the likelihood of finding a solution.

A rule of thumb approach which may be more prone to error but faster in problem solving is a
Back to Cognitive Psych Home page 
A rule of thumb approach which may be more prone to error but faster in problem solving is a
Back to Topics


What problem

Heuristics are mental shortcuts that allow people to solve problems and make judgments quickly and efficiently. These rule-of-thumb strategies shorten decision-making time and allow people to function without constantly stopping to think about their next course of action.

What is heuristic rule

A heuristic is a rule-of-thumb, or a guide toward what behavior is appropriate for a certain situation. Heuristics are also known as “mental shortcuts” (Kahneman, 2011). Such shortcuts can aid us when we face time pressure to decide, or when conditions are complex and our attention is divided.

Which problem

Heuristics are rules of thumb that often, but not always, help us solve problems. They are shortcuts that are faster than algorithms, but they are not always reliable.

What is a heuristic rule?

Heuristic rules are shortcuts that deliver quicker decisions than traditional methods when problem-solving in computing and elsewhere – we could even say they are used to reach educated guesses.