THEORY AND PRACTICE

PROBLEM SOLVING: the

improve reasoning process, he rec-
ommended adopting the problem-
WHAT IS IT? HOW solving method for all subjects and
grade levels. His five steps are:
CAN WE TEACH IT? 1. Becoming aware of difficulty
2. Identifying the problem
BY ALLAN C. ORNSTEIN 3. Assembling and classifying data
Since Charles Judd at the University of and formulating hypotheses
4. Accepting or rejecting tentative
Chicago and Edward Thorndike at
Columbia University showed that learn- hypotheses
5. Formulating conclusions and
ing could be explained in terms of gen- 1
eral principles of thinking and methods evaluating them.’
of attacking problems transferred to dif- Dewey’s work, as are many problem-
ferent situations, educators and psy- solving models today, is based on ad-
vice. For example, Bradford and Stein
chologists have identified various meth-
ods to teach students how to problem outline the IDEAL method for problem-
solve. solving : Identify the problem; Define it;
John Dewey’s five steps in &dquo;reflec- Explore possible strategies; Act on the
tive thinking&dquo; was a classic model for strategies; and Look at the effects of
your efforts.22
problem solving until the 1950s when
Many successful problem solvers ad-
Piaget’s work and more sophisticated
models employing various cognitive and vocate heuristic approach; that is, they
a
look at the main problem and then re-
information-processing strategies were
introduced in the United States. Never- duce the main problem or its complexity
into components.
theless, Dewey’s model is still con-
sidered practical today, especially For Newell and Simon, the method
for dealing with a problem is twofold:
among science and math teachers, al-
though cognition theorists consider it first, construct a representation of the
oversimplified. problem, called the &dquo;problem space,&dquo;
and second, work out a solution that
Dewey’s five steps utilized the natu-
ral involves a search through the problem
sciences, called the &dquo;scientific
method,&dquo; to systematically interpret space.
While attempting to represent or un-
everyday experiences through the rea-
soning process. Since one of the chief 1. John Dewey, How We Think
functions of school, for Dewey, was to (Lexington,
Mass.: D.C. Heath, 1910).
ALLAN C. ORNSTEIN is professor in the 2. John Bransford and Barry S. Stein, The
School of Education, Loyola University of IDEAL Problem Solver (San Francisco:
Chicago. Freeman, 1985).
Downloaded from bul.sagepub.com at UCSF LIBRARY & CKM on April 25, 2015 113
114

derstand the problem, the problem lem when they encounter a situation in
solver may break down the problem, which they must respond but do not
activate old information (from mem- have the immediate solution. Regardless
ory), or seek new information. If the of the exact steps or methods for solving
solution is successful, the task ends. If the problem, the students need basic
it fails, the person backtracks, side- prerequisite information and relevant
tracks, or redefines the problem or the knowledge to assess the situation and to
method used to solve it. Problem solv- work out a strategy for solving the prob-
ing is not linear; the problem solver may lem.
jump around, skip, or combine steps. The strategies used are related to the
In this connection, Richard Cyert students’ age and the specific problem.
suggests that problem solving is heur- According to researchers, not all suc-
istic and includes 10 steps: cessful students will use the same
0 Keeping the big picture in mind with- strategies, and more than one strategy
out bogging down with details can often be used to solve the same
. Avoiding early commitment to a par- problem. A number of strategies can be
ticular hypothesis when several hypoth- used to solve even a simple problem
eses are possible (e.g., 6 + 6 ?).5 =

. Simplifying the problem by using The effects of success or failure in

phrases, symbols, or formulas problem solving are similar to the ef-
. Changing approaches if one is not fects with other outcomes of learning.
working Repeated failure results in giving up,
0 Asking yourself questions and at- substituting a different goal, or pro-
tempting to answer them ducing undesirable behavior.
. Questioning your own assumptions if Some basic problem-solving strate-
necessary gies do seem to emerge, however. In a
0 Working backwards if needed to classic study, Benjamin Bloom pointed
work out solutions out the differences between successful
0 Keeping in mind partial solutions that and unsuccessful students engaged in
later may be combined problem solving. Although the subjects
~ Using metaphors and analogies were college students, the basic differ-
4
. Talking about the problem.4 ences apply to most students of various

ages who are involved in problem-

Successful/Unsuccessful Problem
solving activities, so long as they have
Solvers. reached the developmental stage of
Students are confronted with a prob- logical thinking, or what Piaget called
&dquo;formal mental operations.&dquo;
3. Allen Newell and Herbert Simon, Human
Problem Solving (Englewood Cliffs, N.J.: 5. Michael Pressley, "The Relevance of the
Prentice-Hall, 1972) Good Strategy User Model to the Teaching
r4. Richard Cyert, "Problem-Solving and of Mathematics," Educational Psychologist
Education Policy," in Problem Solving and Spring 1986, pp. 139-161; David N. Perkins
Education: Issues in Teaching and Re- and Gavriel Saloman, "Teaching for Trans-
search, edited by D. Tuma and R. Reif fer," Educational Leadership, September
(Hillsdale, N.J.: Elbaum, 1980), pp. 3-7. 1988, pp. 22-32.

Downloaded from bul.sagepub.com at UCSF LIBRARY & CKM on April 25, 2015
 115

1. Comprehending the problem. Suc- volving a plan, set of steps, or

cessful problem solvers reacted to se- procedures-as well as the ability to
lected cues and immediately began to evaluate and modify performance.
work out a solution. Unsuccessful stu- The important implication for learn-
dents missed cues and often misinter- ing to solve problems is to possess
preted the problem. metacognitive skills. Based on a review
2. Employing previous knowledge. The of the research, some of these skills or
successful group utilized previous processes that have been found to dis-
knowledge to solve the problem. The tinguish successful problem solvers are:
unsuccessful students possessed the ac- 1. Comprehension monitoring. Know-
cessory information but did not utilize ing when one understands or does not
it. They often did not know where or understand something; evaluating one’s
how to start. performance.
3. Style of problem-solving behavior. 2. Understanding decisions. Under-
Successful students were more active standing what one is doing and why.
and could verbalize what they were do- 3. Planning. Taking time to develop a
ing. They simplified the problem, strategy; considering options; pro-
whenever possible, or broke it down ceeding without impulse.
into parts if they could not deal with the 4. Estimating task difficulty. Estimating
whole. The unsuccessful students were the difficulty of the task; allocating suf-
rarely able to clarify or concisely state ficient time for difficult problems.
what they were doing; they often did not 5. Task presentation. Keeping on target
attempt to analyze the various parts. and staying with the task; being able to
4. Attitude toward problem solving. The ignore internal and external distractions;
successful problem solvers had con- maintaining definite direction in one’s
fidence, whereas the unsuccessful group thinking.
lacked confidence when faced with the 6. Coping strategies. Staying calm and
problem. The first group was motivated being able to cope when things are not
or viewed the problem as a challenge; going smoothly or easily. Not giving up
the second group became frustrated and or becoming anxious or frustrated.

gave up.66 7. Internal cues. Searching for internal

There is general consensus, today, or context clues when confronted with
that metacognitive skills (or processes) difficult or novel tasks, problems, etc.
are transferable competencies that play a 8. Retracking. Backtracking, looking up
significant role in problem solving and definitions, rereading previous
high-order thinking. Metacognitive information-and knowing when to
skills or processes represent knowledge backtrack.
of how to do something-usually in- 9. Noting and correcting. Noting and
correcting; using logical approaches;
6. Benjamin S. Bloom and Lois J. Broder, double checking; being aware of incon-
"Problem Solving Process of College Stu-
dents," in Supplementary Educational sistencies, contradictions, or gaps in
Monograph, No. 73 (Chicago: University of performance.
Chicago Press, 1950). 10. Flexible approaches. Being willing

Downloaded from bul.sagepub.com at UCSF LIBRARY & CKM on April 25, 2015
116

to use alternative approaches; knowing stages of cognitive growth, not until the
when to search for another strategy; child is 11 years old is he or she capable
using random approaches that are sen- of performing many of these meta-
sible and plausible when one’s original cognitive skills (corresponding with
approach has been unsuccessful.’ formal mental operations), and not until
Low-achieving and younger students age 15 is the child capable of fully per-
have less procedural knowledge to per- forming all metacognitive skills effi-
form these metacognitive skills com- ciently.9
pared to high-achieving and older stu- Study Skills, Subjects, Problem
dents.~ The implication for teaching is
that increased knowledge of content or Solving
subject matter does not necessarily pro- Every instructional program incor-
duce changes in metacognitive skills. porates several study skills, especially
These skills develop over years and in as the student passes from grade level to

general reflect high-order thinking pro- grade level. At the elementary school
cesses that cannot be learned or devel- level, study skills are often considered
oped overnight or in one subject. as a separate subject or content area. At

Moreover, age or human growth and the secondary level, study skills are
development is crucial in limiting poten- blended into curriculum as part of the
tial metacognitive skills among stu- subject matter. Every academic subject,
dents. An eight-year-old student is just at the junior high and senior high level,
capable of so much and cannot cogni- incorporates a number of specific study
tively be pushed more than one or two skills, suggested by Figure 1.
as

years beyond his or her developmental Although most educators describe

stage. Corresponding with Piaget’s study skills in terms of reading and lan-
guage skills, study skills for our pur-
poses here have been divided into three
7. Jere E. Brophy, "Research Linking
Teacher Behavior to Student Achievement," categories: verbal, symbol, and inves-
Educational Psychologist, Summer 1988, tigative. Secondary students often can-
pp. 235-286; David F. Lohman, "Predicting not engage in problem solving without
Mathematic Effects in the Teaching of High- facility in study skills. The deployment
Order Thinking Skills," Educational Psy- and utilization of study skills may be
chologist, Summer 1986, pp. 191-208; and considered generic content for problem
Richard K. Wagner and Robert J. Sternberg,
"Alternative Conceptions of Intelligence solving, regardless of the subject.
and Their Implications for Education," Re- Problem-solving properties or catego-
view of Educational Research, Summer ries may be referred to as operations.
1984, pp. 179-224. Figure 1 lists 27 different problem-
8. Brophy, "Research Linking Teacher Be-
havior to Student Achievement"; Lyn Corno
and Richard E. Snow, "Adapting Teaching 9. Jean Piaget, The Origins of Intelligence in
to Individual Differences Among Learners," Children (New York: International Universi-
Handbook of Research in Teaching, pp. ties Press, 1952); Jean Piaget and Barbel
605-629, edited by Wittrock; and Lohman, Inhelder, The Early Growth of Logical
"Predicting Mathemathanic Effects in the Thinking in the Child (London: Routledge
Teaching of High-Order Thinking Skills." Kega Paul, 1964).
Downloaded from bul.sagepub.com at UCSF LIBRARY & CKM on April 25, 2015
 117

Downloaded from bul.sagepub.com at UCSF LIBRARY & CKM on April 25, 2015
118

solving operations which cut across sub- mosphere required.

ject matter and are written in terms of a Such teachers often see problem solv-
hierarchy. No attempt has been made to ing as a chore, not as a commitment,
define the different terms because of and in competition with, not comple-
space limitations; the list is theoretical mentary to, the teaching-learning proc-
and has not been tested for validity. ess. The outcome is that in a typical

Based on Piagetian theory, students classroom at the elementary or secon-

should be able to perform these dary level, problem solving is rarely ob-
problem-solving operations by age served ; in fact, it is one of the least
14 when they enter into the formal observed instructional methods regard-
operations stage of cognitive develop- less of the subject.’°
ment. The operations help teachers and Problem solving is no easy task to
instructional leaders discover relation- teach or learn. But how can students
ships that exist among study skills, sub- improve problem solving in the class-
ject areas, and problem solving. room ? Specific instructional procedures
are outlined by Good and Grouws, who
Procedure for Teaching Problem
have identified five problem-solving
Solving processes that can be used in the
While most teachers give problem teaching-learning process:
solving high marks for importance, 1.Attending to prerequisites. Solving
many have not learned how to incor- new problems is based largely on under-

porate it into their lessons. Others view standing previously learned skills and
problem solving as interfering with or concepts of the subject.
taking time away from curriculum cov- 2. Attending to relationshships. Sub-
erage, since many skills and processes jects comprise a large body of logical
are required. and closely related ideas; the teacher
More important, problem solving should emphasize meaning and interpre-
does not express itself in neatly orga- tation of ideas.
nized packages-that is, easily observed 3. Attending to representation. The
behaviors and outcomes-that can be more the student is able to represent a

simply listed in lesson plans, taught to problem in context with concrete or

students, or evaluated as they unfold. real-world phenomena, the better able
There is a certain amount of vagueness the student is to solve the problem.
and risk-and extra class time is 4. Generalizability of concepts. Teach-
needed-when teachers and students ers must explain the general appli-

engage in problem solving.

Still another reason is that some 10. John I. Goodlad, A Place Called School
teachers do not see problem solving as (New York: McGraw-Hill, 1983); John I.
part of their instructional function and Goodlad and Francis Klein, Behind the
are apt to incorporate this method only
Classroom Door (Worthington, Ohio: Jones
Publishers, 1970); and Report of the 1977
mechanically (i.e,. following Dewey’s National Survey of Science, Mathematics,
five steps) or skeptically, without the and Social Studies Education (Washington,
dynamic and enthusiastic classroom at- D.C.: National Science Foundation, 1977).

Downloaded from bul.sagepub.com at UCSF LIBRARY & CKM on April 25, 2015
 119

cability of the idea to students; skills and are no precise equivalents in other lan-
processes that apply to many settings guages or subjects and whose meanings
should be practiced. are therefore context specific.’2
5. Attending to language. Teachers In general, the improvement of stu-
should use precise terminology of their dents’ thinking-from what Lipman
subject, and students must learn basic calls &dquo;ordinary thinking&dquo; to &dquo;problem
terms and concepts of the subject.&dquo; solving&dquo;-depends on the students’
Although the authors describe the ability to identify good criteria for their
problem-solving process for math- opinions or solutions. The criteria
ematics, the five processes seem to should be based on logic and must be
apply subjects. The authors con-
to all &dquo;relevant,&dquo; &dquo;reliable,&dquo; &dquo;coherent,&dquo;
tend that teachers can learn to incor- and &dquo;consistent&dquo; with acceptable stan-
porate these steps into their own teach- dards in that particular subject or field. 13
ing processes, and students can also In this connection, the most charac-
learn to use these thinking processes. teristic feature of problem solving is that
Matthew Lipman adds five other it discovers its own weakness and recti-
processes that are more subtle and also fies what is at fault in its procedures;
cut across situational contexts and sub- thus, it is self-correcting.
ject matter. Problem-solving teaching
should take into account (1) exceptional Problem-Solving Strategies
or irregular circumstances: certain In general, teachers must provide stu-
things may be permissible under certain dents with good problem-solving strate-
conditions and not so under most condi- gies and practice using such strategies.
tions, or vice versa; (2) special limi- An extensive catalog of instructional
tations or constraints: certain theories, strategies is discussed below.
principles, or concepts may have limi- 1. The problem-solving process in most
tations or contingencies when they don’t cases is either schema driven, with little

apply or when they break down; (3) search for solution procedures because
overall configurations: something may appropriate procedures are activated by
be wrong or objectionable when taken recognizing the particular problem type,
out of context but valid or proper as a or based on search strategies, in which

whole, or vice versa; (4) evidence may the problem solver looks for or decides
be atypical-overgeneralizing from a on a particular strategy before solution
14
small sample or one experiment is procedures are tried or worked out.
somewhat risky; and (5) some meanings
do not translate well from one context to 12. Matthew Lipman, "Critical Thinking—
another-that is, there are terms, ex- What Can It Be?" Educational Leadership,

pressions, concepts, etc. for which there September 1988, pp. 38-43.
13. Ibid. Also see Joan B. Baron and Robert
J. Sternberg, eds., Teaching Thinking Skills
11. Thomas L. Good and Douglas A. (New York: Freeman, 1987).
Grouws, "Increasing Teachers’ Under- 14. Mary L. Glick "Problem-Solving
standing of Mathematical Ideas Through In Strategies," Educational Psychologist, Win-
Service Training," Phi Delta Kappan, June ter/Spring 1986, pp. 99-120; Penelope Pet-
1987, pp. 778-783. erson, "Teachers’ and Students’ Cognitional

Downloaded from bul.sagepub.com at UCSF LIBRARY & CKM on April 25, 2015
120

2. Experts and subject specialists (in- straightforward problems.

cluding teachers) possess schemata rel- examining problems already
evant to problems in their area of exper- worked out results in general
tise or subject. Consequently, they strategies for a similar group of
rarely engage in search strategies as problems; it does not prepare the
novices often do. Experts and specialists problem solver for the atypical
possess schemata organized around pro- problem or a new set of problems.
cedural knowledge about solutions and 4. The use of different problem-solving
their applicability. strategies is influenced by the age and
3. In the absence of appropriate knowledge base of the problem solver
schemata, the successful problem solver and by the problem itself. Students ages
usually proceeds to a search strategy. 14 an older should be able to process
Search strategies include, among others, information and acquire sufficient
a means-ends analysis, analogies, knowledge of a subject to attack prob-
matching, sequencing, and examining lems in selected subjects like &dquo;little&dquo;
problems already worked out. 15 experts subject specialists.
or

. Means-ends strategies are goal 5. As students gain more knowledge in a

based and require that the problem subject area, they learn more search
solver eliminate differences be- strategies and are able to make more
tween the current and the goal con- moves and shifts in problem solving.

dition ; they usually lead to memory 6. In analyzing a problem, students

overload. must understand what they have to do to
. Analogies are important for mak- obtain a solution: where to begin, how
ing the unfamiliar problem more to move forward, how to backtrack,
familiar; novice problem solvers how to combine strategies or subparts of
are likely to engage in superficial problems, and when they have a satis-
analogies, as opposed to those factory answer or solution.
based on solution procedures. 7. Knowledge of when, where, and how
. Matching involves selecting a plan to apply certain strategies is also crucial;
of action and carrying it out ac- students must learn to apply specific
cordingly ; the process of matching strategies. Students should be taught
plans and actions continues until that often the result of apply-
errors are

the problem is solved. ing the incorrect strategies to problems

*sequenceing strategies (step-by- rather than lack of ability or effort.
step) is an integral part of training 8. Although some problem strategies cut
and can be used for solving across several subjects, others rest on
the adequacy of a knowledge base that
Knowledge for Classroom Teaching and supports the strategy. Without a knowl-
Learning," Educational Research, Ju- edge base in a given subject, it is nearly
ne/July 1988, pp. 5-14. impossible to solve complex problems.
15. Glick, "Problem-Solving Strategies";
9. An extensive knowledge base permits
Pressley, "The Relevance of the Good both skipping and mixing strategies to
Strategy User Model to the Teaching of
Mathematics." solve problems.

Downloaded from bul.sagepub.com at UCSF LIBRARY & CKM on April 25, 2015
 121

10. In some advanced subjects, knowl- ence, generalize from findings, move
edge when to apply certain strategies
of from concrete to abstract and then from
cannot be taught to students. Students abstract to abstract. In all cases, there is
must induce the knowledge indirectly or need to stress objectivity of judgment in
intuitively and in other cases by exam- light of existing conditions.
ining problems already worked out, 13. Students can learn to use specific
means-ends analysis, etc.
16 strategies through consistent practice in
11. Perhaps the most fundamental a variety of settings so eventually their

problem-solving skill is learning to strategies develop into broad schemata.

monitor or diagnose progress when little Diverse opportunities to use strategies
is being achieved, when mistakes are increase abstraction and strategy knowl-
being made, etc. Students must reflect edge ; abstraction of the above
on what they are doing and know when 17 broadens one’s ability to
strategies
to modify or change strategies. cope with different problems.
12. When solving problems, students 14. Abstraction of strategies is less
must be taught to compare and contrast, likely with low-achieving and younger
relate parts to the whole, decompose the students. More explicit teaching is nec-
whole to parts, control variables, under- essary for these students.
stand cause-effect relationships, recog- 15. Students must learn that time and
nize inconsistencies in statements or effort are important for problem solv-
findings, distinguish between bias and ing. Time and effort should be matched
objective data, generalize from experi- with task situations; students should be
cognizant of making good use of time
16. James G. Greeno, "Trends in the and sensitive when they are wasting
Theory of Knowledge for Problem- time because of distractions or lack of
Solving," in Problem Solving and Educa- focus.
tion, edited by D. Tuma and F. Reif (Hill-
sdale, N.J.: Erlbaum, 1980), pp. 9-23; 17. David N. Perkins and Gavriel Salomon
Matthew W. Lewis and John R. Anderson,
"Teaching for Transfer," Educational
"Discrimination of Operator Schemata in Leadership, September 1988, pp. 22-32;
Problem Solving: Learning from Exam-
Pressley, "The Relevance of the Good
ples," Cognitive Psychology, January 1985, Strategy User Model to the Teaching of
pp. 26-65. Mathematics. "

This Month. From page v.

From these activities in school reform, state officials and to provide them with the
we can see how important it is for school kind of information that school administra-
administrators to communicate with their tors know best.

Downloaded from bul.sagepub.com at UCSF LIBRARY & CKM on April 25, 2015