Types of students' justifications

Sowder, Larry; Harel, Guershon . The Mathematics Teacher ; Reston  Tomo 91, N.º 8,  (Nov 1998): 670-
675.

Enlace de documentos de ProQuest

RESUMEN (ABSTRACT)
 
Students often have a difficult time articulating their justifications for arriving at mathematical conclusions.
Sowder and Harel offer a framework for thinking about students' justifications in an effort to help shape their
mathematical reasoning.

TEXTO COMPLETO
 
The comments above, from the inarticulate "I just know it" to the sophisticated reference to a proof, present
different ways that students might justify particular mathematical results. Our aim in this article is to give a
framework for thinking about students' justifications, with an eye toward shaping their mathematical reasoning.
The framework has arisen from our teaching and our interviews with secondary school and college students, but it
also relects the work of others (e.g., Chazan [1993]; Goetting [1995]; Yerushalmy [1993]). A more elaborate version
of the framework appears in Harel and Sowder (1998).
Mathematical proofs are perhaps the ultimate in justifications, but the traditional approaches to them have been
unsuccessful for many students, including college students as well as those at the high school level. When we
have asked college mathematics majors about their experience with proof, too often they have responded with
such comments as, "What the words `mathematical proof' mean to me is a necessary evil" or "I remember hating
those proofs in high school geometry. I remember thinking, `Why do I need to prove this? It's so obvious"' or "For
what reason? I guess I really never felt they were that important, since I was willing to trust whatever brilliant
mathematician thought up the theorem." Research studies of secondary students' proof understanding and
performance (e.g., Senk [1985]; Silver and Carpenter [1989]; Thompson [1996]; Tinto [1988]; Williams [1980])
substantiate that in the past we have not been very successful with this aspect of the curriculum.
Standard 3, Mathematics as Reasoning, of the NCTM's Curriculum and Evaluation Standards for School
Mathematics (1989) recognizes the importance of students' reasoning. The standard states that we cannot
continue to settle for results like many that we obtain from the traditional curriculum. Hence, if the Standards
document's call for student justifications in all areas of mathematics, rather than just in geometry, results in
students' learning that reasoning in mathematics is as important as the facts of mathematics, then we will have
made progress. If the Standards document's support for activities in which students generate conjectures leads
naturally to questions of justification, then we have set a stage for mathematical reasoning.
The framework in this article uses the notion of a student's proof scheme. Proving, or justifying, a result involves
ascertaining-that is, convincing oneself-and persuading, that is, convincing others. An individual's proof scheme
consists of whatever constitutes ascertaining and persuading for that person. Arguments by the same person but
in different contexts or at different times might fall into different categories, and a specific line of reasoning might
involve a combination of proof schemes. Hence, even though some schemes are clearly more sophisticated than
others, the categories of proof schemes described in this article should not be considered as a hierarchy in which a
given person always operates at the "highest" level. The word proof in proof scheme is used in the broader,
psychological sense of justification rather than in the narrower sense of mathematical proof.

PDF GENERADO POR SEARCH.PROQUEST.COM Page 1 of 7

THE PROOF SCHEMES
Proof, or justification, schemes can be organized into three categories: externally based proof schemes, empirical
proof schemes, and analytic proof schemes, with subcategories for each (see fig. 1). In illustrating and elaborating
on these proof schemes, we also point out shortcomings that we have noticed from our interviews with students.
Externally based proof schemes
In the externally based proof schemes, both what convinces the student and what the student would offer to
persuade others reside in some outside source. This source might be an authority, as in the authoritarian proof
scheme; the form of the argument, as in the ritual proof scheme; or meaningless manipulations of symbols, that is,
the symbolic proof scheme. Let us examine each of these types of proof schemes.
The authoritarian proof scheme. Students demonstrate the authoritarian proof scheme when they believe that they
can rely only on a textbook, a teacher statement, or perhaps a confirmation from a more knowledgeable classmate
to justify a result. Referring to an authority is not entirely bad, of course. Even experts in mathematics may accept
another person's result without examining the details of the reasoning that led to the result. But when students
have recourse only to such sources for convincing themselves or others, cause for concern exists. A learned
helplessness-"I have to see whether I can find it in the book" or "I don't remember"-or even a rejection of the value
of proofs-"I skip the proofs, since it's the results that are important"-are the worst effects of a sole reliance on
authority. These responses may be the consequences of instruction that emphasizes only authority-given results
without paying attention to the reasoning that led to the results.
As the classroom authority, the teacher certainly has important roles in setting tasks, guiding discussions, raising
doubts, pointing students in directions of inquiry, giving explanations, and stating results. But in the long run, it is
counterproductive to teach procedures, say, for operations with fractions or with algebraic expressions, without
supporting reasoning or the expectation of reasoning and sense making from the students. For example,
multiplication of two binomials can be taught by rote, as with FOIL, that is, sum the product of the First terms, then
the Outer terms, then the Inner terms, and finally the Last terms. Examining why FOIL works not only sets a tone of
expectation that the routine classroom practice will involve explanations but also indicates that the explanations
may form a basis for handling such similar calculations as the product of a binomial and a trinomial. Explanations
may also lead to a deeper understanding of such related topics as the algorithm for the product of two two-digit
numbers. Many current uses of small groups-for example, on tasks requiring interpretations and open to
alternative reasonable approaches-not only provide opportunities for student discussion and cooperation but also
remove the teacher as the sole focus and the source of all knowledge. The ritual proof scheme. A student who
judges the correctness of an argument solely by the form of the argument rather than by the correctness of the
reasoning involved is exhibiting the ritual proof scheme (Martin and Harel 1989). For example, college students
from traditional curricula may accept as proofs only reasoning that is arranged in the twocolumn format that they
learned in high school geometry. Even when they offer sound arguments, they may doubt that they have given a
proof because their argument is not organized in such a format or because it may not involve enough
mathematical notation or computation. The Standards document's recommendations that paragraph proofs be a
greater part of the curriculum, that two-column proofs not be the only forms labeled as proofs, and that proof be a
concern in the nongeometry part of the curriculum all should help us combat the ritual proof scheme.
These suggestions do not mean that two-column proofs should be discontinued. Good instructional reasons may
exist for organizing a proof in that form. Rather, the use of the word proof exclusively in settings that involve two-
column explanations should be avoided.
The symbolic proof scheme. One side of symbolic proof schemes is bad, whereas the other is good. When
students treat the symbols as though they have a life independent of any meaning or of any relationship to the
quantities in the situation in which they arose, we are seeing the bad side, the external side. This type of symbolic
reasoning scheme allows one to react to or to evaluate 25 x 23 with 415, to replace sin (x + y) with sin x + sin y, or
to replace a with a + b. Once again, an expectation that assertions are subject to justifications can set a tone that
might help discourage mindless symbol pushing.

PDF GENERADO POR SEARCH.PROQUEST.COM Page 2 of 7

The good side of symbolic reasoning lies in the well-known power of symbols, especially in algebra. In solving a
system of linear equations for a story problem, for example, one does not have to attach to each step in the
process a meaning in terms of the context of the problem. An advantage-or disadvantage-of coordinate geometry
over synthetic or coordinate-free versions is that algebra does "everything" and one does not have to do as much
reasoning. In many cases a knowledgeable person can "unpack" correct symbolic reasoning; such a person has a
"transformational" proof scheme, to be discussed later.
Empirical proof schemes
Justifications made solely on the basis of examples show another type of proof scheme. Psychologists point out
that most natural human concept formation is indeed based on examples-and sometimes on a specific example
(Medin 1989). Even in this article, most readers probably appreciate the examples, either as a means of
understanding the point or as a means of checking their understanding. Our work suggests two types of empirical
proof schemes, the perceptual proof scheme and the examplesbased proof scheme.
The perceptual proof scheme. Geometry teachers are familiar with students who have arrived at conclusions on
the basis of their perceptions of a single, or occasionally several, drawings. Students who also try to convince
others by showing them a drawing are operating under the perceptual proof scheme. For example, in proving that
segments connecting the midpoints of the sides of an isosceles trapezoid form a rhombus, one college student
consistently drew a square as her isosceles trapezoid and reasoned using the right angles she saw in that
drawing. Selected optical illusions can show the danger of making conclusions on the basis of a drawing (fig. 2).
Technology offers help, inasmuch as the dynamic feature of many geometry-software packages allows drawings
to be altered quickly. Hence, it is relatively easy to examine whether an argument continues to apply to these
altered drawings. Geometry work within small groups can help identify students who are reasoning perceptually
and can alert such students to the need for other types of justifications. For example, the question "Is Rosanna's
quadrilateral a rectangle?" arose naturally in response to Rosanna's ruler-compass method for constructing a
rectangle (fig. 3), with some students reacting perceptually but the questioning student clearly calling for another
sort of justification.
The examples-based proof scheme. Convincing oneself, and others, by evaluating a conjecture through one or
more examples uses the examples-based proof scheme. Problems that invite an examplesbased approach, with
the resulting patterns breaking down, can alert students to the dangers of such reasoning. Examples of such
problems, usable at a variety of grade levels, follow.
Problem 2:
How many different lengths are on a geoboard with ten rows and ten pegs in each row?
Problem 3:
Does n^sup 2^ - 79n + 1601, for n a whole number, always give a prime number?
Problem 4:
Is 1 + 1141n^sup 2^ ever a perfect square, for some positive integer n?
Problem 5:
Should 0^sup 0^ be defined according to the pattern from 0^sup 4^ = 0, 0^sup 3^ = 0, 0^sup 2^=0,0^sup 1^= 0 or the
pattern from 4^sup 0^ =1, 3^sup 0^=1, 2^sup 0^ =1, 1^sup 0^ =1?
To avoid a lengthy digression from proof schemes, let us make just these comments about the problems. In
problem 1, the attractive pattern breaks down at the target calculation. Similarly, in problem 2, an approach based
on smaller geoboards, and perhaps supported by a convincing drawing, suggests a pattern that does not give the
correct number-19-for the case of six rows with six pegs in each row, as well as for subsequent cases. Problem 3
may be well known; for n = 0 through 79, the expression does give a prime number, but it does not for n = 80.
Leonard (1997) uses the less formidable n2 - n + 41. Problem 4, from Davis (1981), is most dramatic: the
expression is not a perfect square for integers from 1 through 30 693 385 322 765 657197 397 207, but it is a
perfect square for the next integer! This problem certainly shows that patterns should not be trusted, even those
that are based on more than 30 septillion cases, but it also illustrates the power of mathematical reasoning. No

PDF GENERADO POR SEARCH.PROQUEST.COM Page 3 of 7

one has actually checked all those cases, but mathematical reasoning assures the results given. Problem 5 shows
that using patterns to justify a result can lead to ambiguities. Such problems notwithstanding, reasoning based on
empirical evidence has a positive side. Mathematicians value examples and nonexamples highly for the added
illumination they can give to an idea. Noticing a particular feature in a drawing and seeing a pattern in a series of
examples are indeed excellent sources of conjectures. Examining several examples is a sensible approach to the
task "Find a shortcut for squaring a number ending in 5" (fig. 4). Similarly, Friedlander and Hershkowitz (1997)
illustrate the use of pattern searching in solving certain algebra problems; the searches lead to conjectured
generalizations that are then subject to justification.
In mathematics, however, our students must become aware of the tentative nature of conjectures formed on the
basis of examples, and they must grow in their abilities to offer additional and better justifications than those
offered by examples. Students should encounter problems like those presented previously, to experience that
patterns alone can lead them astray. To leave the impression that patterns alone are always completely adequate
justifications of conjectures can suppress better ways of thinking. For example, in trying to predict the number of
edges in an n-gonal prism, one might examine triangular prisms, quadrilateral prisms, pentagonal prisms, and so
on, and notice, that is, conjecture, from the pattern that the number of edges of an n-gonal prism is 3n. Although
that approach is commendable, it need not be the stopping point. An argument that depends on "seeing" the 3n
edges, for example, could follow such patternbased work: Each of the two bases of the prisms have n edges, plus
the n lateral edges, giving 3n edges in all. Finding a pattern is one thing, but seeing why a generalization holds is
another.
Analytic proof schemes
Mathematics teachers probably regard the analytic proof schemes as giving the ultimate types of justifications in
mathematics. As figure 1 indicates, we see two subtypes, transformational analytic proof schemes and axiomatic
analytic proof schemes.
The transformational proof schemes. The general feature of these schemes is that students' justifications are
concerned with the general aspects of a situation and involve reasoning oriented toward settling the conjecture in
general. For example, the argument presented previously for the number of edges on an n-gon would show a
transformational proof scheme if one perceives the underlying structure behind the pattern. The argument is
general and involves reasoning rather than the counting and pattern finding of an examples-based proof scheme.
For the squaring-a-number-ending-in-5 task of figure 4, a transformational proof scheme might lead to reasoning
based on (lOn + 5)2. As an important side benefit of that reasoning, one realizes why a similar shortcut does not
work for squaring numbers ending in a digit other than 5, and that the shortcut will also work for products like 26 x
24 and 77 x 73, although this latter shortcut will likely require prompting. We regard the transformational proof
scheme as a necessary precedent to the last proof scheme, the axiomatic proof scheme.
The axiomatic proof schemes. In mathematics, a body of knowledge can be organized so that subsequent results
are logical consequences of preceding ones. A careful organization involves undefined terms, definitions,
assumptions, and theorems. Present-day Euclidean geometry is a familiar example, with "point" and "line" being
undefined and statements like "Two points determine a unique line" being assumed. A student who is comfortable
working with such a system is showing an axiomatic proof scheme. The Standards document identifies local
axiomatics, the study of a short sequence of definitions and theorems, as a venue for examining the organizational
role of proof and seeing the possibility of different organizations for the same body of material.
"Leaping," as it were, into instruction expecting an axiomatic proof scheme without having had previous, and
significant, experiences with justifications is expecting a lot from a student. In Fawcett's seminal study (1938),
geometry students first spent a considerable amount of time dealing with the difficulties of making acceptable
definitions for reallife topics before developing what turned out to be an axiomatic treatment of geometry. Healy's
(1993) similar informal experiments along the same lines should also result in axiomatic proof schemes for many
students.
CLOSING REMARKS

PDF GENERADO POR SEARCH.PROQUEST.COM Page 4 of 7

The proof schemes outlined in this article give us a way to evaluate the justifications given by our students, so that
we can plan instruction to move them toward more sophisticated ways of reasoning. A student who relies on
authority might profit from small-group work in which conjectures are to be formed and in which no authority
exists; conjectures are to be settled within the group, without appeal to the teacher. Students who place full faith
in examples should at least find out that examples can betray them and that patterns found in several examples
are not completely trustworthy, as in the problems presented in this article.
On the one hand, we are dismayed by the results of studies that suggest that students in traditional curricula are
not interested in learning the "why" after learning the "how" (e.g., Hiebert and Wearne [1988]; Kamii [1994]). On the
other hand, we are encouraged by studies that show that even primarygrade children can thrive in environments in
which reason giving by the students is a natural and expected part of the mathematics lessons (e.g., Campbell and
Robles [1997]; Carpenter et al. [1998]; Cobb et al. [1991]; Kamii [1994]). It seems clear that to delay exposure to
reason giving until the secondary-school geometry course and to then expect an instant appreciation for more
sophisticated mathematical justifications is an unreasonable expectation. To make "mathematics as reasoning"
an obvious concern, we must continuously expose our students throughout their schooling in mathematics to the
value of arguments that focus on the "why" of results and not just on the results themselves.
References
REFERENCES
References
Campbell, Patricia F., and Josepha Robles. "Project IMPACT: Increasing the Mathematical Power of All Children
and Teachers." In Reflecting on Our Work: NSF Teacher Enhancement in K-6 Mathematics, edited by Susan N. Friel
and George W. Bright, 179-86. Lanham, Md.: University Press of America, 1997.
References
Carpenter, Thomas P., Megan L. Franke, Victoria R. Jacobs, Elizabeth Fennema, and Susan B. Empson. "A
Longitudinal Study of Invention and Understanding in Children's Multidigit Addition and Subtraction." Journal for
Research in Mathematics Education 29 (January 1998): 3-20. Chazan, Daniel. "High School Geometry Students'
Justification for Their Views of Empirical Evidence and Mathematical Proof." Educational Studies in Mathematics
24 (1993): 359-87.
References
Cobb, Paul, Terry Wood, Erna Yackel, John Nicholls, Grayson Wheatley, Beatriz Trigatti, and Marcella Perlwitz.
"Assessment of a Problem-Centered SecondGrade Mathematics Project." Journal for Research in Mathematics
Education 22 (January 1991): 3-29. Davis, Philip J. "Are There Coincidences in Mathematics?"American
Mathematical Monthly 88 (1981): 311-20.
Fawcett, Harold P. The Nature of Proof. Thirteenth Yearbook of the National Council of Teachers of Mathematics
(NCTM), New York: Teachers College, Columbia University, 1938; Reston, Va.: NCTM, 1995.
References
Friedlander, Alex, and Rina Hershkowitz. "Reasoning with Algebra." Mathematics Teacher 90 (September 1997):
442-47.
Goetting, Mary M. "The College Student's Understanding of Mathematical Proof." Ph.D. diss., University of
Maryland, 1995.
Harel, Guershon, and Larry Sowder. "Students' Proof Schemes." In Research in College Mathematics Education III,
edited by Alan H. Schoenfeld, Jim Kaput, and Ed Dubinsky, 234-83. Providence, R.I.: American Mathematical
Society, 1998.
Healy, Christopher C. Build-a-Book Geometry. Berkeley, Calif.: Key Curriculum Press, 1993.
References
Hiebert, James, and Diana Wearne. aA Cognitive Approach to Meaningful Mathematics Instruction: Testing a Local
Theory Using Decimal Numbers." Journal for Research in Mathematics Education 19 (November 1988): 371-84.
Kamii, Constance. Young Children Continue to Reinvent Arithmetic. New York: Teachers College Press, 1994.

PDF GENERADO POR SEARCH.PROQUEST.COM Page 5 of 7

References
Leonard, Bill. "Proof: The Power of Persuasion." Mathematics Teacher 90 (March 1997): 202-5. Martin, W. Gary, and
Guershon Harel. "Proof Frames of Preservice Elementary Teachers." Journal for Research in Mathematics
Education 20 (January 1989): 41-51.
References
Medin, Douglas L. "Concepts and Conceptual Structure." American Psychologist 44 ( 1989): 1469-81. National
Council of Teachers of Mathematics (NCTM). Curriculum and Evaluation Standards for School Mathematics.
Reston, Va.: NCTM, 1989. Senk, Sharon L. "How Well Do Students Write Geometry Proofs?" Mathematics Teacher
78 (February 1985): 448-56.
References
Silver, Edward A., and Thomas P. Carpenter. "Mathematical Methods." In Results from the Fourth Mathematics
Assessment, edited by Mary M. Lindquist, 10-18. Reston, Va.: National Council of Teachers of Mathematics, 1989.
Thompson, Denisse R. "Learning and Teaching Indirect Proof." Mathematics Teacher 89 (September 1996): 474-82.

References
Tinto, Patricia P. "Students' Views on Learning Proof in High School Geometry." Paper presented at the meeting of
the American Educational Research Association, New Orleans, April 1988. Williams, Edgar. "An Investigation of
Senior High School Students' Understanding of the Nature of Mathematical Proof." Journal for Research in
Mathematics Education 11 (May 1980): 165-66. Yerushalmy, Michal. "Generalization in Geometry." In The
Geometric Supposer: What Is It a Case Of? edited by Judah L. Schwartz, Michal Yerushalmy, and Beth Wilson, 57-
84. Hillsdale, N.J.: Lawrence Erlbaum Associates, 1993.
AuthorAffiliation
Larry Sowder, [email protected], taught high school mathematics in Indiana and is presently teaching
at San Diego State University, San Diego, CA 92182-7720. He is interested in students' proof writing and, in general,
in their problem-solving performance. Guershon Harel, [email protected], is a professor at Purdue University
and associate editor of the American Mathematical Monthly. His research centers on the epistemology of
mathematics and its application in mathematics curricula.

DETALLES

Materia: Mathematics education; Cognition &reasoning

Título: Types of students' justifications

Autor: Sowder, Larry; Harel, Guershon

Título de publicación: The Mathematics Teacher; Reston

Tomo: 91

Número: 8

Páginas: 670-675

Año de publicación: 1998

Fecha de publicación: Nov 1998

PDF GENERADO POR SEARCH.PROQUEST.COM Page 6 of 7

 Editorial: National Council of Teachers of Mathematics

Lugar de publicación: Reston

País de publicación: United States, Reston

Materia de publicación: Education--Teaching Methods And Curriculum, Mathematics

ISSN: 00255769

CODEN: MATAAP

Tipo de fuente: Scholarly Journals

Idioma de la publicación: English

Tipo de documento: Feature

Número de acceso: 04014441

ID del documento de 204626041

ProQuest:

URL del documento: https://search.proquest.com/docview/204626041?accountid=15293

Copyright: Copyright National Council of Teachers of Mathematics Nov 1998

Última actualización: 2016-10-15

Base de datos: Materials Science &Engineering Collection

ENLACES
Linking Service

Copyright de la base de datos  2019 ProQuest LLC. Reservados todos los derechos.

Términos y condiciones Contactar con ProQuest

PDF GENERADO POR SEARCH.PROQUEST.COM Page 7 of 7