Musings on Factoring of Polynomials
The factoring of polynomials (1) is a topic in algebra
about which students have vastly different feelings:
• for some, factoring is a prime example of a technique for which the most appropriate response is, "When
are we ever gonna use this?.
• for some others, factoring provides an enjoyable puzzle, akin, perhaps, to diagramming sentences in an English
I should be explicit about the difference between factoring a polynomial, such as x2 - 5x +6 =(x - 2)(x - 3) and
factoring an integer, such as 629 = 17 · 37.
|In the first example, the equals sign means that,
for every x, the value
of the polynomial, x2 – 5x – 6, for any number x, equals the product
of the numbers, x – 2 and x – 3, for that number x, and so there are
infinitely many numerical equalities contained in the single algebraic
equation; while, in the second example, we are told that the number,
629, equals (or can be expressed as, or can be factored into) the
product of the two numbers, 17 and 37.
Few students see any significance in factoring, and many teachers
join students in disdain or distaste for factoring. Sandie
Coelho, PIMMS’ former Associate Director for Mathematics,
showed me a sheaf of questions and answers about what’s factoring
good for? or why teach factoring?—the teacher-discussants
were seriously trying to find justification for their teaching of a
topic that they feared has no significance. In that sheaf, not one
person could provide a reason for learning how to factor polynomials!
Indeed, one frustrated respondent said that, if asked by a
student, "What good is it to know how to factor x2 – 5x – 6?",
he would answer, "Factor this expression or I’ll pull the trigger of this loaded pistol that I’m holding to your temple." This
mean Dickensian teacher isn’t much worse than the sarcastic Euclidean mentor who tells his assistant to give a dime to the
student "since he needs to have some reason to study."
If I am asked, "Should students in a high school algebra class spend substantial time to learn factoring of polynomials?",
I respond with a clear-cut "NO and YES." Let me explain:
If the student’s experience with factoring embodies no more than a semi-infinite number of problems like the following:
"By factoring, solve the equation 168 + 15x – 18x2 = 0,"
then the student has been cheated.
It is desirable, of course, for the student to realize that
•168 + 15x – 18x2 =
3(56 + 5x – 6x2);
•168 + 15x – 18x2 = 0 56 + 5x – 6x2 =0; that is, the two quadratic equations have the same solutions and
therefore are equivalent;
• For real p, q, p· q = 0 p = 0, or q = 0, or both;
• If, for all real numbers, x, 56 + 5x – 6x2 = (a + bx) (c + dx) for integers a, b, c, d,
then ac =56 and bd = – 6; and
• There are only finitely many values of a, b, c, d that need be tested to find the possible factors.
Many students do not appreciate any of the statements contained in the foregoing bullets. Consider the amusing
blunder made by some students in applying bullet #3:
Asked to solve 56 + 5x – 6x2 =0 , the student (correctly) rewrites the equation as: 6x2 – 5x = 56,
permitting the factorization: x(6x – 5) = 56.
Now the student applies a home-grown (and false) generalization of our "fundamental principle" (bullet #3) to
x = 56 or 6x – 5 = 56 (!)
The student failed to recognize that the statement of bullet #3 is true if and only if the constant is zero.
To teach factoring as a method for solving quadratic equations may be considered an exercise in futility, because
the factorable quadratic expressions form so insignificant a subset of all quadratic expressions.
But, "completing the square" to obtain the "quadratic formula"—now there’s something worthwhile! I say this with
enthusiasm because the process of completing the square leads to reducing the quadratic equation whose roots we seek to
a couple of linear equations whose roots we already know how to find—in analogy with the sort of reduction used by Tart-
aglia and Ferrari to solve cubic and quartic equations (2). Often the solutions (answers) of a problem are almost unrelated to
the value of solving the problem. The important values of solving classroom problems are the procedures (patterns, or sequences
of steps) that are recognized as useful also in solving numerous other problems.
In response to the query, "When are we ever gonna use this?", the student probably expects an example that seems
practical, from his or her point of view, and solving an algebraic equation does not strike the average ninth-grader as a really
urgent matter. In calculus, factoring is useful to break a rational fraction expression into partial fractions, permitting us to
integrate the original expression and to deal with some infinite series. But, for a high school teacher to appeal to these applications
as reasons for learning to factor forces the teacher to say, "Trust me, you’ll need this sometime in the future"—a
pedagogically weak argument (3).
Thus, I say NO! to spending lots of time in elementary and intermediate algebra classes on factoring quadratic expressions
in order to solve quadratic equations.
On the other hand, I think that it is not only desirable but really pedagogically necessary to work through some factoring to
get across the important concepts contained in the five bullets above. This explains part of the reason for the YES in my answer
to the factoring question.
Now let me strengthen my YES by outlining three problems, not involving the solution of equations, whose solutions
require that students have an understanding of factoring and some facility with the process. The first of these problems
is appropriate for a first course in algebra, while the second and third may require more mathematical maturity than typical
1 ∙ 3 + 1 = 4,
2 ∙ 4 + 1 = 9,
3 ∙ 5 + 1 = 16,
4 ∙ 6 + 1 = 25,
5 ∙ 7 + 1 = 36.
What pattern do you notice? Express your guess (conjecture) in words. How do you know whether
this pattern continues indefinitely?
The pattern might be described as follows: The product of two consecutive odd numbers, or of two consecutive even numbers,
plus 1, equals a perfect square - the square of the “skipped” number.
A direct algebraic argument that this pattern is valid in general follows from letting the first number in the product be
called n. It doesn’t matter whether it is even or odd. Then the second number in that product will be n + 2; the product
of the two numbers is n(n + 2); and the left hand side of the equations of the problems will be n(n + 2) + 1.
But n(n + 2) + 1 = n2 + 2n + 1, which equals (n + 1)2.
|This is clearly the square of (n + 1), the number
between n and (n + 2).
The equality n(n + 2) + 1 = (n + 1)2 is an identity, valid for
all integers, n, and thus validates the guess expressed in
Prob. 2 (Given to a Middletown, CT sixth-grade class by
the cosmopolitan mathematician and educator, W. W.
Sawyer, over 50 years ago. The problem also appears as the second exercise in The Theory of Numbers, a text written by the
noted American mathematician, Robert D. Carmichael, about 100 years ago.)
1 ∙ 2 ∙ 3 ∙ 4 + 1 = 25,
2 ∙ 3 ∙ 4 ∙ 5 + 1 = 121,
3 ∙ 4 ∙ 5 ∙ 6 + 1 = 361, and
4 ∙ 5 ∙ 6 ∙ 7 + 1 = 841.
What pattern(s) do you notice? Express your guesses (conjectures) in words.