Teaching For Thinking Research Paper

1. Introduction

1.1 Definitions

Teaching for thinking refers to instruction that is intended to improve the effectiveness of a people’s thinking. This definition consists of two parts: (a) teaching—teaching for thinking involves instruction, and (b) thinking—the goal of the instruction is to improve someone’s ability to solve problems. Teaching refers to manipulations to a learner’s environment which are intended to foster a change in the knowledge of the learner, whereas thinking refers to cognitive activity aimed at solving a problem. In sum, teaching for thinking involves instructional manipulations that are intended to foster the learner’s knowledge of how to solve problems.

Teaching for thinking is a type of cognitive strategy instruction (Halpern 1992, Mayer and Wittrock 1996, Pressley and Woloshyn 1995). In cognitive strategy instruction learners are taught how to control their cognitive processing such as how to think, how to learn, how to remember, or how to motivate oneself. Thus, cognitive strategy instruction involves fostering metacognitive knowledge—knowledge of how one’s mind works—and strategic knowledge—knowledge of how to develop plans for accomplishing goals—rather than procedural knowledge—knowledge of how to carry out a routine—and declarative knowledge— factual knowledge about some topic.

1.2 Rationale

Students are expected to be effective thinkers but rarely are taught how to think. Students are expected to be able to solve problems but rarely are given productive experience that would foster the development of problem-solving skills. Although the formal curriculum in any subject area may emphasize the learner’s acquisition of content knowledge, an often neglected part of the curriculum concerns the learner’s being able to use what was learned to solve problems. Instruction in how to think is sometimes not part of the formal curriculum but rather is part of the hidden curriculum; although unstated it is expected that learners will be able to use what they have learned in problem solving. For these reasons, there has been a growing call for explicit instruction in how to be a more effective problem solver (Halpern 1992, Mayer and Wittrock 1996, Pressley and Woloshyn 1995).

1.3 Example

The Gestalt psychologist, Wertheimer (1959) offers a classic example of the challenges of teaching for thinking. He describes a classroom in which a teacher presents a lesson on how to find the area of a parallelogram. The teacher draws a parallelogram on the chalk board, drops a perpendicular line from the top base to the bottom base (which is called the height), measures the height, measures the base, multiplies height times base and gives the result as the area (using the formula area height base). After the demonstration, the teacher calls on a student to solve a similar problem at the board and the student is successful in carrying out each of the steps. Later the students do very well on a quiz in which they must compute the area of parallelograms.

In spite of the apparent success of the lesson, Wertheimer (1959, p. 15) criticized the teaching for its lack of emphasis on thinking: ‘Most people would say, ‘‘This is an excellent class; the teaching goal has been reached.’’ But observing the class, I feel uneasy, I am troubled. ‘‘What have they learned,’’ I ask myself. ‘‘Have they done any thinking at all? Have they grasped the issue? Maybe all they have done is little more than blind repetition.’’ ’ As a test, Wertheimer presents an unusually shaped parallelogram, and the students respond by saying ‘We haven’t had this yet.’

In contrast to teaching for rote learning, Wertheimer suggests that students should be encouraged to discover that the triangle on one of the parallelograms could be cut off and placed on the other end to form a rectangle. In this way, students could see that a parallelogram is simply a rectangle in disguise. Since they already know how to find the area of a rectangle, the solution to parallelogram problems now becomes much easier. Wertheimer referred to this method of teaching as learning by understanding, and claimed that it resulted in superior problem-solving transfer, that is, in being able to use what was learned in new situations. According to Wertheimer, rote teaching methods lead to reproductive thinking—being able to repeat what was taught—whereas meaningful teaching methods can lead to productive thinking—being able to use what was taught in new situations.

When the goal is teaching for thinking—that is the promotion of productive thinking—instructors need to consider four issues: what to teach, how to teach, where to teach, and when to teach. In addressing these four issues, Mayer (1997) has compared four commonsense principles of teaching for thinking with corresponding alternative principles based on research in cognitive science.

2. What To Teach

The first issue concerns what to teach. Should instruction focus on thinking as a single monolithic ability or as a collection of smaller component skills? The common-sense view casts thinking as a single ability: the ability to solve problems is a unitary skill so teaching for thinking involves improving one’s mind in general through learning certain school subjects that require mental discipline and orderly thinking. An implication of the single ability view of what to teach is that the mind is a sort of mental muscle that needs to be exercised to increase its overall functioning. When the mind is improved through learning a school subject that exercises the mental muscle, then the learner will also show improvements in a wide variety of academic tasks.

In the early 1900s the dominant view among educational scholars was the doctrine of formal discipline, namely, the idea that learning certain school subjects, such as Latin, geometry, and logic, promoted proper habits of mind including mental discipline and orderly thinking. One of the first tasks of the newly established field of educational psychology was to test the doctrine of formal discipline in a scientifically rigorous way. For example, in a convincing series of studies, Thorndike and co-workers found that students who had learned Latin performed no better than students who had not learned Latin in learning a new school subject such as bookkeeping (Thorndike 1924). Similarly, more recent claims that learning programming languages such as Logo would improve children’s minds were also put to the test; researchers found that students who learned Logo did not outperform non-Logo learners on problem-solving tasks unrelated to Logo (Mayer 1988). Overall, the research evidence failed to support the common-sense view of thinking as a single ability.

In contrast, a common theme in modern research in cognitive science is that a person’s performance on any cognitive task—such as solving a word problem or writing an essay—depends on the degree to which the person possesses relevant component skills (Sternberg 1990). To write an essay, for example, students need to engage in three main cognitive processes—planning, translating, and reviewing (Flower and Hayes 1980); to solve an arithmetic story problem students need to engage in four cognitive processes—translating each sentence into a mental representation, integrating the sentences into a coherent mental model of the situation, planning a set of steps to solve the problem, and executing the solution plan (Mayer 1992).

According to the multiple skills view, teaching for thinking should focus on helping students learn the collection of component skills that are relevant for each important cognitive task rather than on improving the mind overall. For example, Pressley and Woloshyn (1995) have shown how focused instruction in component skills can improve students’ performance in reading, writing, maths, and other cognitive tasks. In answer to the question of what to teach, their answer is to teach a collection of smaller component skills.

3. How To Teach

The second issue concerns how to teach. Should instruction focus on encouraging students to generate correct answers (i.e., the product of problem solving) or on the methods used to solve problems (i.e., the process of problem solving)? The common-sense view favors the product-of-problem-solving approach: as in learning any skill, the best approach to learning of thinking skills is to get plenty of practice in answering questions along with feedback on whether or not the answers are correct. Although drill-and-practice instructional methods can be effective in improving performance on problems like those presented during instruction, they sometimes fail to help students solve transfer problems that go beyond those presented during instruction.

For example, in a classic study, Brownell and Moser (1949) taught students to solve two-column subtraction problems (such as 76 – 48 = ____) by drilling them in the steps of the procedure or by showing how the steps correspond to using sticks that bundled into groups of 10 (i.e., by using concrete manipulatives). Although students who learned with concrete manipulatives did not outperform the traditionally taught students on retention problems—two column subtraction problems like those given during instruction—they did excel on transfer problems, such as solving three-column subtraction problems. Brownell and Moser concluded that students understand the material better when the emphasis changes from product to process.

Apprenticeship approaches offer a straightforward way to focus on problem-solving processes because learners attempt to model the cognitive processes of more skilled performers. For example, in an early study, Bloom and Broder (1950) taught unsuccessful students how to answer exam questions by modeling the problem-solving processes of successful students. First, a successful student described his or her thought process while answering an exam question. Second, the unsuccessful student described his or her thought process while answering the same question, and then the unsuccessful student compared his or her process with that of the successful student. After 10 to 12 sessions with different questions, the unsuccessful students were much better able to answer new exam questions than were a group of matched students who received no training. Bloom and Broder concluded that teaching for thinking is most effective when instruction focuses on the process of problem solving rather than solely on the product of problem solving.

According to the process-of-problem-solving view, teaching for thinking should focus on helping students learn the methods for how to solve problems rather than solely on practice in getting the right answer. Modern research in cognitive science has demonstrated the value of apprenticeship approaches that emphasize the process of problem solving (Collins et al. 1989). Successful techniques include reciprocal teaching of reading comprehension skills (Brown and Palinscar 1989). In answer to the question of how to teach, their answer is to model the process of problem solving.

4. Where To Teach

A third issue concerns where to teach. Should instruction be given as a general, independent, standalone course or should instruction in thinking be incorporated into specific subject matter areas? The common sense view favors the independent approach: all students should be required to take a course in problem solving, in addition to their regular content courses. An implication is that problem-solving skills are best learned as an isolated subject, separated from the context in which they will be applied.

The Productive Thinking Program was intended as a stand-alone course in problem solving, aimed at helping elementary school students learn to generate and test hypotheses (Covington et al. 1966). It consisted of a series of 15 cartoon-like booklets, each describing a mystery or detective story. The reader was asked to participate in solving the case by generating and testing hypotheses, along with help from various characters in the booklet who modeled appropriate processes for generating and testing hypotheses. Students who took the course tended to outperform matched students who had not taken the course on solving problems like those in the booklets, but not on problems that were from a different context (Mayer 1992). Thus, in spite of attempts to teach context-free problem-solving skills, students tended to perform best in using the skills within the context they were presented. These results tend to refute the commonsense view that teaching for thinking should occur as a separate course.

In contrast, cognitive science research points to the domain-specificity of learning. For example, people were asked to determine which of two products was the best buy, such as a 10-ounce can of peanuts at 90 cents or a 4-ounce can at 45 cents. In school, the most common approach is to compute unit cost, yielding 9 cents per ounce versus 11.25 cents per ounce. However, in a supermarket, people tended to use a ratio approach in which they reasoned that the larger one costs twice as much but contains more than twice as many ounces. Similarly, Nunes et al. (1993) found that students who could solve arithmetic problems without error in the context of their job as street vendors made many errors when confronted with the same problems in a school setting. In addition, research on expertise demonstrates that people who are expert problem-solvers in one domain (such as physics or chess) are generally not superior in problem solving in other domains (Mayer 1992). These results support the idea that the usability of problem-solving strategies depend on the context in which they are learned.

According to the specific-domain view, teaching for thinking should take place within a specific context rather than in a context-independent environment. Research on the lack of transfer of school-learned mathematics (Nunes et al. 1993) is consistent with the specific-domain view. In answer to the question of where to teach, their answer is to teach cognitive skills within the specific context they will be applied.

5. When To Teach

The final issue concerns when to teach. Should instruction focus on teaching higher-order thinking skills only after the learner has mastered lower-level basic skills or should instruction on thinking begin when the learner is still a novice? The common-sense view favors the prior-mastery approach: students cannot understand higher-order skills until they have mastered lower-level ones.

However, cognitive science research on teaching reading, writing, and mathematics shows that students can benefit from learning higher-order thinking skills before they have completely mastered the basics. In reading, for example, elementary school students can benefit from learning how to summarize a passage, how to predict what will come next, and how to recognize portions that need clarification even if they are still not fluent in basic decoding skills (Brown and Palinscar 1989). In writing, students can benefit from writing creative essays even before they have mastered spelling, grammar, and punctuation (Kellogg 1994). In mathematics, students can benefit from direct instruction in how to represent word problems even before they have completely automatized their computational fluency (Sternberg and Ben-Zeev 1996). In contrast to the prior-mastery view, this research provides support for an approach that can be called constraint removal—providing scaffolding support that enables students to work on tasks they have not yet mastered (such as allowing a student to dictate an essay).

According to the constraint-removal approach, teaching for thinking can begin while the learner is still a novice. In answer to the question of when to teach, supporters of this approach argue that it makes no sense to wait until the student has automated all lower level skills.

6. Summary

In summary, teaching for thinking involves helping students to be able to use what they learn to solve new problems. In contrast to common-sense views, teaching for thinking is likely to be most successful when the focus is on (a) teaching thinking as a collection of smaller component skills rather than as a single monolithic ability, (b) teaching the process of how to solve problems rather than solely the product of problem solving, (c) teaching thinking within a specific subject domain rather than as an isolated domain-free course, and (d) teaching while students are novices rather than waiting until they have mastered all basic skills.

Bibliography:

1. Bloom B S, Broder L J 1950 Problem-solving Processes of College Students. University of Chicago Press, Chicago
2. Brown A L, Palinscar A S 1989 Guided, cooperative learning and individual knowledge acquisition. In: Resnick L B (ed.) Knowing, Learning, and Instruction. Erlbaum, Mahwah, NJ, pp. 393–451
3. Brownell W A, Moser H E 1949 Meaningful versus Mechanical Learning: A Study in Grade 3 Arithmetic. Duke Uni ersity Research Series in Education (No. 8). Duke University Press, Durham, NC
4. Collins A, Brown A L, Newman S E 1989 Cognitive apprenticeship: Teaching the crafts of reading, writing, and mathematics. In: Resnick L B (ed.) Knowing, Learning, and Instruction. Erlbaum, Mahwah, NJ, pp. 453–94
5. Covington M V, Crutchfield R S, Davies L B 1966 The Productive Thinking Program. Brazelton, Berkeley, CA
6. Flower L S, Hayes J R 1980 Identifying the organization of writing processes. In: Gregg L W, Steinberg E R (eds.) Cognitive Processes in Writing. Erlbaum, Mahwah, NJ, pp. 3–30
7. Halpern D F (ed.) 1992 Enhancing Thinking Skills in the Sciences and Mathematics. Erlbaum, Mahwah, NJ
8. Kellogg R T 1994 The Psychology of Writing. Oxford University Press, New York
9. Mayer R E (ed.) 1988 Teaching and Learning Computer Programming. Erlbaum, Mahwah, NJ
10. Mayer R E 1992 Thinking, Problem Solving, Cognition, 2nd edn. Freeman, New York
11. Mayer R E 1997 Incorporating problem solving into secondary school curricula. In: Phye G D (ed.) Handbook of Academic Learning. Academic Press, San Diego, pp. 473–92
12. Mayer R E, Wittrock M C 1996 Problem-solving transfer. In: Berliner D C, Calfee R C (eds.) Handbook of Educational Psychology. Macmillan, New York, pp. 47–62
13. Nunes T, Schliemann A D, Carraher D W 1993 Street Mathematics and School Mathematics. Cambridge University Press, Cambridge, UK
14. Pressley M, Woloshyn V 1995 Cognitive Strategy Instruction. Brookline Books, Cambridge, MA
15. Sternberg R J 1990 Metaphors of Mind. Cambridge University Press, Cambridge, UK
16. Sternberg R J, Ben-Zeev T (eds.) 1996 The Nature of Mathematical Thinking. Erlbaum, Mahwah, NJ
17. Thorndike E L 1924 Mental discipline in high school studies. Journal of Educational Psychology 15: 1–22
18. Wertheimer M 1959 Productive Thinking. Harper and Row, New York