The four-integer search for factoring ax squared plus bx plus c, the special patterns when a and c are perfect squares, pulling out a common monomial first, solving equations that must be rewritten in standard form, and maximising revenue through the average of the zeros.
Subject: Algebra 2 · 74 slides · symbolic lesson
Open the interactive version of this deck
Title
Algebra 2 · Lesson 4.4
The trial and error has an end, and you can see it from the start
Objectives
Lesson 4.3 factored trinomials whose leading coefficient was 1. There you only had to split the constant, because the square term left you no choice about the two x terms. This lesson removes that comfort: the leading coefficient must be split as well.
That is the lesson people describe as endless guessing. It is not endless, and the whole purpose of the next hour is to show you where the end is. There is a finite list of candidates, you can write it down, and reading two signs off the trinomial deletes half of it before you multiply anything.
Two habits run through every slide. Always look for a common factor before anything else, because it shrinks the numbers you have to search. And always multiply your answer back out, because expanding is quick and it is the only thing that can catch a lost sign.
McDougal Littell Algebra 2 (Texas Edition), Ch. 4 Quadratic Functions and Factoring — Lesson 4.4 Solve ax^2 + bx + c = 0 by Factoring §4.4, p. 259 — Before, Now, Why and the Key Concept box
Warm-up
Start with the lesson you have just finished, so you can see exactly what changes. Write your answer down before reading on, because the comparison is only worth something if you committed to a method first.
\[ x^2 - 7x + 12 = 0 \]
Discussion prompt
Factor the left-hand side and solve the equation. Then say, in one sentence, what you did not have to decide.
Check both answers in the original. Three squared is nine, and nine minus twenty-one plus twelve is zero. Four squared is sixteen, and sixteen minus twenty-eight plus twelve is zero as well. Both work, which is what two factors promised.
Now change one number and watch the comfort disappear. Replace the leading 1 by a 5 and you get five x squared minus seventeen x plus six, and suddenly the two x terms are a choice too: the 5 can go into either bracket, and the six can be split in several ways beside it.
The zero product property does not change at all. Once the left-hand side is a product of two brackets equal to zero, one of them has to be zero, exactly as here. Everything new in this lesson is about reaching that product in the first place.
McDougal Littell Algebra 2 (Texas Edition), Ch. 4 Quadratic Functions and Factoring — Lesson 4.4 Solve ax^2 + bx + c = 0 by Factoring §4.4, p. 259 — Before and Now
Section
Part 1
Concept
Here is the difference drawn rather than described. On the left is a trinomial with a leading coefficient of 1; on the right the same shape with a 5 in front. The boxes show every product that has to come out right.
Figure (svg): Two multiplication boxes side by side. In the first, for x squared plus seven x plus twelve, both sides begin with a plain x. In the second, for five x squared minus seventeen x plus six, one side begins with 5x and the other with x.
Look at the top-left cell of each box. On the left it must be x times x, and there is no other way to reach x squared with integer coefficients, so the two leading terms are decided for you. On the right the 5 has to come from somewhere, and it has to be split between the two brackets.
So a trinomial with a leading coefficient of one asks for two numbers, and this lesson asks for four. That is the honest size of the change, and everything else follows from it.
The bottom-right cell is the same in both boxes: it is always the constant term. And the two cells on the other diagonal are the only ones that carry an x on its own, so they are the only ones that decide the middle term.
McDougal Littell Algebra 2 (Texas Edition), Ch. 4 Quadratic Functions and Factoring — Lesson 4.4 Solve ax^2 + bx + c = 0 by Factoring §4.4, p. 259 — the Key Concept box
Notation
The book states the whole method as one identity. It looks forbidding and it is actually a promise: expand the right-hand side and you get to read off exactly what the four integers have to do.
Annotate
On: \( ax^2 + bx + c = (kx + m)(lx + n) \)
Read those four notes as a plan and not as a definition. Two of them hand you finite lists, the third tells you what to test, and the fourth deletes half the possibilities before you start.
Notice which of the four coefficients is easy and which is hard. The first and last coefficients each come from a single product, so they are simply factor pairs. Only the middle one mixes the two lists together, and so only the middle one has to be searched.
McDougal Littell Algebra 2 (Texas Edition), Ch. 4 Quadratic Functions and Factoring — Lesson 4.4 Solve ax^2 + bx + c = 0 by Factoring §4.4, p. 259 — the Key Concept box and the Factor Expressions note
Concept
This is the sentence to carry away from Part 1. If the middle coefficient were a product you could factor it and be done. It is a sum of two products, and that is why arranging the numbers matters as much as choosing them.
Figure (svg): The product of k x plus m and l x plus n, with an arc joining k x to n labelled k times n and a second arc joining m to l x labelled l times m, and the three resulting coefficients listed at the right.
Trace the two arcs. The outer one multiplies the leading coefficient of the first bracket by the constant of the second; the inner one does the opposite pairing. Those two results are the only x terms in the expansion, so together they are the middle term.
\[ (kx + m)(lx + n) = klx^2 + (kn + lm)x + mn \]
That also explains something that puzzles people. Swapping m and n between the brackets leaves the product of the constants alone and leaves the product of the leading coefficients alone, but it changes which number meets which, so the middle coefficient changes. The order is part of the answer.
So the search is not really over four numbers. It is over arrangements: a way of pairing one list against the other. Counting the arrangements is what makes the search finite.
McDougal Littell Algebra 2 (Texas Edition), Ch. 4 Quadratic Functions and Factoring — Lesson 4.4 Solve ax^2 + bx + c = 0 by Factoring §4.4, p. 259 — Example 1, the arrangement table
Concept
Before you write down a single candidate, read two signs off the trinomial. This is the cheapest work in the lesson and it removes more possibilities than anything else you will do.
Figure (svg): A decision tree. The top box asks for the sign of c. One branch leads to m and n having the same sign, which then splits by the sign of b into both positive and both negative; the other branch leads to m and n having opposite signs.
The reasoning behind each branch is the sign rule for multiplication, nothing more. The product of the two constants has to equal c, so if c is positive the two constants agree in sign, and if c is negative they disagree.
When they agree, the sign of the middle coefficient settles which sign it is. Two negatives multiply to a positive constant but add to something negative, so a negative middle coefficient with a positive constant means both are negative.
\[ 5x^2 - 17x + 6: \quad c > 0 \text{ and } b < 0 \;\Longrightarrow\; m < 0 \text{ and } n < 0 \]
Apply that to the lesson's first example and you have already halved the work. The constant is a positive six, so the two numbers agree; the middle coefficient is negative, so they are both negative. Every arrangement with a plus in it is gone.
McDougal Littell Algebra 2 (Texas Edition), Ch. 4 Quadratic Functions and Factoring — Lesson 4.4 Solve ax^2 + bx + c = 0 by Factoring §4.4, p. 259 — Example 1, the sign argument
Sorting
Practise the sign reading on six trinomials, with no factoring at all. For each one ask what c does first, and only then look at what b does. Your answer should take about three seconds.
Sort into buckets
Sort each trinomial by the signs its two constant terms must have.
Notice that the constant term does all the deciding about agreement, and the middle coefficient only becomes relevant once you know the two numbers agree. Reading them in that order keeps the rule short.
The opposite-signs case is the one that still leaves work. Knowing the signs disagree does not tell you which of the two candidates is the negative one, so those trinomials keep twice as many arrangements as the others. That is the case Example 2 is built for.
McDougal Littell Algebra 2 (Texas Edition), Ch. 4 Quadratic Functions and Factoring — Lesson 4.4 Solve ax^2 + bx + c = 0 by Factoring §4.4, p. 263 — Exercises 3 to 11
Concept
This is the slide that answers the complaint. Trial and error feels endless because nobody writes the trial down. Once you do, you can count it, and the count is small.
Take five x squared minus seventeen x plus six. Factor the leading coefficient and factor the constant, keeping the leading pair positive as the book does.
| what to split | the pairs available | how many |
|---|---|---|
| the 5 | 5 and 1 | one pair |
| the 6, both negative | −1 and −6; −2 and −3; −3 and −2; −6 and −1 | four arrangements |
| altogether | one leading pair against four constant arrangements | four candidates |
Four. Not endless, four, and each one takes about five seconds to expand. If you write the four rows down and work along them you will finish this problem in under a minute and you will know you have finished, which guessing never tells you.
The two orderings of the same pair both appear because the order matters, exactly as the arrow picture showed. Negative two with negative three is a different candidate from negative three with negative two, since the 5 meets a different number in each.
The count grows when the leading coefficient has more factor pairs, and that is precisely why the next part takes any common factor out first. Shrinking the leading coefficient shrinks this table.
McDougal Littell Algebra 2 (Texas Edition), Ch. 4 Quadratic Functions and Factoring — Lesson 4.4 Solve ax^2 + bx + c = 0 by Factoring §4.4, p. 259 — Example 1
Worked example
The book's first example, worked as a written search rather than a guess. Notice that the table is the method: the point of writing it is that you can see the end of it.
\[ 5x^2 - 17x + 6 \]
Factor the leading coefficient and the constant
Why: five is prime, so the only positive pair is 5 and 1; six can be split as 1 and 6 or as 2 and 3.
\[ a = 5 = 5 \cdot 1, \qquad c = 6 = 1 \cdot 6 = 2 \cdot 3 \]
Read the signs
Why: the constant is positive, so the two numbers agree in sign; the middle coefficient is negative, so both are negative.
\[ mn > 0 \text{ and } b < 0 \;\Longrightarrow\; m < 0, \; n < 0 \]
That is the halving, applied. Without it you would have eight candidates to test; with it you have four, and you have not multiplied anything yet.
Write out the four candidates and expand each
Why: with k equal to 5 and l equal to 1, the middle coefficient is five times n, plus m.
\[ (5x + m)(x + n) = 5x^2 + (5n + m)x + mn \]
| m and n | the candidate | what it expands to |
|---|---|---|
| −6 and −1 | (5x − 6)(x − 1) | 5x² − 11x + 6 |
| −1 and −6 | (5x − 1)(x − 6) | 5x² − 31x + 6 |
| −3 and −2 | (5x − 3)(x − 2) | 5x² − 13x + 6 |
| −2 and −3 | (5x − 2)(x − 3) | 5x² − 17x + 6 |
Every row has the right first and last term, because every row uses factors of 5 and factors of 6. Only the middle column separates them, which is the point Part 1 has been making.
Take the row whose middle term matches
Why: negative two with negative three is the arrangement that produces negative seventeen x.
\[ 5x^2 - 17x + 6 = (5x - 2)(x - 3) \]
Check by multiplying back out
Why: five x times x is five x squared; the two middle products are negative fifteen x and negative two x, which add to negative seventeen x; and negative two times negative three is six.
\[ (5x - 2)(x - 3) = 5x^2 - 15x - 2x + 6 = 5x^2 - 17x + 6 \]
Expanding is not optional politeness. It is a ten-second calculation that catches every lost sign, and a lost sign is the most common error in this whole lesson.
McDougal Littell Algebra 2 (Texas Edition), Ch. 4 Quadratic Functions and Factoring — Lesson 4.4 Solve ax^2 + bx + c = 0 by Factoring §4.4, p. 259 — Example 1
Picture it
The table told you which row wins. The box tells you why the middle term comes out as it does, and it is the picture to reach for whenever a factorisation looks wrong and you cannot see where.
Figure (svg): A multiplication box for 5x minus 2 times x minus 3. The four cells read 5x squared, minus 15x, minus 2x and 6, with the two middle cells picked out in yellow.
Read the box as a multiplication table. The row headings are the terms of the first bracket, the column headings are the terms of the second, and each cell is the product of its row and its column. Four cells, four products, and their sum is the trinomial.
The two yellow cells are the only ones carrying a single x, so only they contribute to the middle term. Negative fifteen x and negative two x add to negative seventeen x, and that is the match the table was hunting for.
One honest caution. Once a sign is negative this is no longer a picture of an area, because no rectangle has a side of length negative two. It is a bookkeeping grid that keeps every product in its own place, and that is exactly what it is good for.
Draw the grid whenever the signs are getting away from you. Filling four cells is harder to get wrong than expanding along a line, because each cell only involves two things.
Prediction
Before working through the second example, get a feel for how the arrangement controls the middle coefficient. All four candidates below have the same first and last terms; only the pairing differs.
Predict first
Of these four candidates for 5x squared minus something x plus 6, which produces the middle coefficient furthest from zero?
Correct: (5x - 1)(x - 6)
Why: The middle coefficient is five times the second bracket's constant, plus the first bracket's constant. Putting the largest constant, six, where the five can multiply it gives thirty, and only one is added to that, for a middle coefficient of negative thirty-one. Pair the six with the one instead and the five only gets to multiply a small number.
The useful habit hiding in that question is a size estimate. Before testing anything, ask roughly how big a middle coefficient you need, then start with the arrangement that gets you near it instead of at the top of the list.
\[ (5x - 1)(x - 6) = 5x^2 - 31x + 6 \]
For negative seventeen you want something in the middle of the range, so the arrangements pairing the 5 with two or with three are the sensible places to begin. That is not guessing; it is ordering a list you have already written.
This also tells you when to give up quickly. If you need a middle coefficient of forty and the largest your lists can reach is thirty-one, the trinomial does not factor over the integers, and you know it after one multiplication.
Worked example
Now the constant is negative, which changes the sign work and leaves twice as many arrangements. Watch how the list is built rather than searched blindly.
\[ 3x^2 + 20x - 7 \]
Factor the leading coefficient and the constant
Why: three is prime and seven is prime, so each list has exactly one pair, which is the smallest possible search.
\[ a = 3 = 3 \cdot 1, \qquad |c| = 7 = 7 \cdot 1 \]
Read the signs
Why: the constant is negative, so the product of the two numbers is negative and they must have opposite signs.
\[ mn < 0 \;\Longrightarrow\; m \text{ and } n \text{ have opposite signs} \]
Opposite signs is weaker information than the last example gave you, so nothing gets deleted here. You keep all four arrangements, which is still a list you can finish.
Write out the four candidates
Why: with k equal to 3 and l equal to 1, the middle coefficient is three times n, plus m.
\[ (3x + m)(x + n) = 3x^2 + (3n + m)x - 7 \]
| m and n | the candidate | what it expands to |
|---|---|---|
| 7 and −1 | (3x + 7)(x − 1) | 3x² + 4x − 7 |
| −1 and 7 | (3x − 1)(x + 7) | 3x² + 20x − 7 |
| −7 and 1 | (3x − 7)(x + 1) | 3x² − 4x − 7 |
| 1 and −7 | (3x + 1)(x − 7) | 3x² − 20x − 7 |
Look down the last column. The four middle coefficients are plus four, plus twenty, minus four and minus twenty: the same two sizes, with both signs. That symmetry is worth noticing, because it means a sign error simply lands you on another row of your own table.
Take the matching row
Why: negative one with seven gives plus twenty x, which is what was asked for.
\[ 3x^2 + 20x - 7 = (3x - 1)(x + 7) \]
Check by multiplying back out
Why: three x times x is three x squared; twenty-one x and negative x add to twenty x; and negative one times seven is negative seven.
\[ (3x - 1)(x + 7) = 3x^2 + 21x - x - 7 = 3x^2 + 20x - 7 \]
Take a moment over the middle step. The two products have opposite signs here, so they partly cancel, and it is the larger that sets the sign of the answer.
McDougal Littell Algebra 2 (Texas Edition), Ch. 4 Quadratic Functions and Factoring — Lesson 4.4 Solve ax^2 + bx + c = 0 by Factoring §4.4, p. 259 — Example 2
Picture it
Draw the second example's box and compare it with the first. The change of sign in the constant changes the arithmetic of the middle term in a way that is easy to see and easy to forget.
Figure (svg): A multiplication box for 3x minus 1 times x plus 7. The cells read 3x squared, 21x, minus x and minus 7, with the two middle cells picked out.
In the first example both yellow cells were negative, so they piled up: fifteen and two became seventeen. Here one is positive and one negative, so they partly cancel: twenty-one and one become twenty.
That is why small factors can produce a surprisingly large middle coefficient when the constant is negative. Three and seven are tiny numbers, and they reach twenty because nothing is cancelling much.
It also explains why the opposite-signs case has more candidates worth testing. With agreement you can predict the size of the sum fairly well; with disagreement the difference can be almost anything, so the list earns its keep.
When you are working on paper, fill in the two corner cells first, since they are forced, and then try arrangements in the two middle cells. You will be changing two entries rather than rewriting the whole thing.
Fill the middle
Your turn, with the list-building done in your head. Seven is prime and three is prime, so each list has one pair, and the negative constant tells you the two numbers disagree in sign.
Fill in the blanks
7x^2 - 20x - 3 = (7x + 1)(x - 3)
Why: With seven and one as the leading pair, the middle coefficient is seven times the second constant plus the first. You need negative twenty, so pair the seven with the three and give the three the minus sign: seven times negative three is negative twenty-one, plus one, which is negative twenty. Expanding confirms it: seven x squared minus twenty-one x plus x minus three.
Notice how the size estimate did the work. Seven times three is twenty-one, which is within one of the twenty you need, so that pairing is the only one worth trying first. The other pairing would give seven times one, plus or minus three, which cannot reach twenty.
The plus sign inside the first bracket is the part people drop. The constant term of the trinomial is negative, so one bracket carries a plus and one a minus, and it is the smaller number that keeps the plus here.
Check it before moving on. Seven x times x is seven x squared, the two middle products are negative twenty-one x and plus x, and one times negative three is negative three. All three terms match.
McDougal Littell Algebra 2 (Texas Edition), Ch. 4 Quadratic Functions and Factoring — Lesson 4.4 Solve ax^2 + bx + c = 0 by Factoring §4.4, p. 260 — Guided Practice 1
Worked example
Every example so far had a prime leading coefficient, which kept one list down to a single pair. Four is not prime, so this is the first genuine two-list search, and it is worth doing slowly.
\[ 4x^2 - 9x + 2 \]
Factor the leading coefficient
Why: four can be split as four and one, or as two and two, so there are two leading pairs to carry through.
\[ a = 4 = 4 \cdot 1 = 2 \cdot 2 \]
Factor the constant and read the signs
Why: two is prime, and the constant is positive with a negative middle coefficient, so both numbers are negative.
\[ c = 2 = (-1)(-2), \qquad m < 0, \; n < 0 \]
Count the candidates before testing any. Two leading pairs, and two ways round to place negative one and negative two, gives four rows. Again a list you can exhaust rather than a search you can only abandon.
Try the leading pair four and one, with the two where the four can reach it
Why: four times two is eight, and eight plus one is nine, which is the size you need.
\[ (4x - 1)(x - 2) = 4x^2 - 8x - x + 2 = 4x^2 - 9x + 2 \]
That was the first row tried, and it worked, because the size estimate pointed at it. Had it failed, the remaining three rows were already written down waiting.
Notice what the other leading pair would give
Why: two and two can only make four x squared with a middle coefficient of negative six, however the constants are arranged, so it cannot reach negative nine.
\[ (2x - 1)(2x - 2) = 4x^2 - 6x + 2 \]
Check the answer by multiplying it out
Why: the two middle products are negative eight x and negative x, which add to negative nine x, and negative one times negative two is two.
\[ (4x - 1)(x - 2) = 4x^2 - 9x + 2 \]
Notice also that the rejected candidate has a bracket, two x minus two, with a common factor of two still inside it. That is a sign of a factorisation that is not in lowest terms, and Part 3 explains why it should have been taken out first.
McDougal Littell Algebra 2 (Texas Edition), Ch. 4 Quadratic Functions and Factoring — Lesson 4.4 Solve ax^2 + bx + c = 0 by Factoring §4.4, p. 260 — Guided Practice 6
Counterexample
Some trinomials do not factor over the integers, and the written list is what lets you say so with confidence instead of suspicion. Here is one, and the point is that you can prove it in four lines.
\[ 2w^2 + w + 3 \]
Discussion prompt
Write out every candidate for this trinomial and show that none of them works. Then say what you would have to change about the middle coefficient to make it factor.
That is a complete argument, not a shrug. Because the list is finite and you wrote all of it, exhausting it is a proof rather than a failure of imagination.
There is a faster test you will meet in Lesson 4.8: the quantity b squared minus four a c. Here it is one minus twenty-four, which is negative, so the equation has no real solutions at all, let alone integer factors. For now the exhausted list is the honest tool you have.
Make a habit of saying which of the two it is when a factorisation will not come. Either you have not found the row yet, or the row does not exist, and only a written list distinguishes those two situations.
McDougal Littell Algebra 2 (Texas Edition), Ch. 4 Quadratic Functions and Factoring — Lesson 4.4 Solve ax^2 + bx + c = 0 by Factoring §4.4, p. 260 — Guided Practice 3
Matching
Four trinomials from the exercise set, four factorisations. You can do this without any searching at all: pick a trinomial, glance at the constant terms on offer, and expand only the one candidate whose signs and sizes could possibly fit.
Match the pairs
Why: Two x squared plus five x plus three has everything positive, so both constants are positive. Eleven z squared plus two z minus nine has a negative constant, and eleven times one minus nine gives exactly two. Fifteen x squared minus two x minus eight is the only one whose leading coefficient splits as five and three. Nine d squared minus thirteen d minus ten needs a leading nine against a lone d, and five with negative two gives negative eighteen plus five, which is negative thirteen.
The one to look at carefully is the trinomial whose leading coefficient is fifteen, because fifteen splits as fifteen and one or as five and three, and the factorisation on offer uses the second split. That is the case where a written list saves the most time.
Notice how much you can settle from the constant term alone. Two of these four trinomials have a negative constant, so their brackets must carry one plus and one minus, and that immediately rules out two of the four candidates for each.
As always, finish by expanding the pairs you chose. Matching by shape is fast and it is also the easiest place to be confidently wrong, so the expansion is what turns a guess into an answer.
McDougal Littell Algebra 2 (Texas Edition), Ch. 4 Quadratic Functions and Factoring — Lesson 4.4 Solve ax^2 + bx + c = 0 by Factoring §4.4, p. 263 — Exercises 3, 7, 8 and 11
Trap
Factoring five x squared minus thirteen x plus six, a very common line:
\[ (5x - 2)(x - 3) \]
The right numbers are all there. Five and one multiply to the leading five, and negative two and negative three multiply to the constant six, so both end terms are correct.
And yet the middle term is wrong, because the 5 has been paired with the 3 instead of the 2.
Expand and the mismatch appears at once.
\[ 5x^2 - 15x - 2x + 6 \]
That is negative seventeen x, not negative thirteen. The correct pairing sends the 5 to the 2.
\[ (5x - 3)(x - 2) \]
Five times two is ten, plus three, which is thirteen. Same four integers, different arrangement, different trinomial.
The guard against this is to write the arrangement table rather than holding it in your head. Both rows appear in it, one line apart, and you pick the one whose middle column matches.
Section
Part 2
Concept
There is a second route through these problems that turns the search into a single question about one pair of numbers. It is usually called the ac-method, and it deserves a justification rather than a recipe, because used blindly it is just another thing to forget.
Figure (svg): A multiplication box with 5x squared in the top left, 6 in the bottom right and question marks in the other two cells, with both diagonals drawn dashed and a note that the two diagonal products must agree.
Look at the two dashed diagonals. The corner cells are the leading term and the constant, so their product is a c times x squared. The other two cells are the two middle pieces, and their product uses the same four integers, so it is the same thing.
\[ (klx^2)(mn) = (kn x)(lm x) = ac\,x^2 \]
So the two pieces of the middle term multiply to a times c and, by definition, add to b. That is one pair of numbers with two known conditions, which is exactly the problem Lesson 4.3 taught you to solve when the leading coefficient was one.
For five x squared minus seventeen x plus six, a times c is thirty, so you want two numbers with product thirty and sum negative seventeen. Negative fifteen and negative two do it, and nothing else does.
McDougal Littell Algebra 2 (Texas Edition), Ch. 4 Quadratic Functions and Factoring — Lesson 4.4 Solve ax^2 + bx + c = 0 by Factoring §4.4, p. 259 — the Key Concept identity, from which this follows
Worked example
The same trinomial as before, along the second route. Compare the two afterwards: the search has moved from four candidates to one pair of numbers, but the numbers are bigger.
\[ 5x^2 - 17x + 6 \]
Multiply the leading coefficient by the constant
Why: five times six is thirty, and thirty is the product the two middle pieces must have.
\[ ac = 5 \cdot 6 = 30 \]
Find two integers with product thirty and sum negative seventeen
Why: the sum is negative and the product positive, so both are negative; the pairs available are one and thirty, two and fifteen, three and ten, five and six.
\[ (-15)(-2) = 30, \qquad (-15) + (-2) = -17 \]
Notice that this list is finite too, and short. There are only four ways to write thirty as a product of two positive integers, so there are only four sums to check, and one of them is seventeen.
Rewrite the middle term as those two pieces
Why: nothing has changed in value, only in appearance: negative fifteen x minus two x is negative seventeen x.
\[ 5x^2 - 15x - 2x + 6 \]
Group the four terms in pairs and take out each pair's common factor
Why: five x comes out of the first pair and negative two out of the second, and both leave the same bracket behind.
\[ 5x(x - 3) - 2(x - 3) \]
That matching bracket is not luck. It happens exactly because the two numbers were chosen to have product a c, which is what the diagonal argument proved a moment ago.
Take out the shared bracket
Why: the expression is now a common factor of x minus three, multiplied by what is left of each term.
\[ (x - 3)(5x - 2) \]
Check by expanding, and against the earlier answer
Why: five x squared minus fifteen x minus two x plus six is the original, and this is the same factorisation Example 1 found by listing arrangements.
\[ (x - 3)(5x - 2) = 5x^2 - 15x - 2x + 6 = 5x^2 - 17x + 6 \]
Two routes, one answer, which is a reassuring thing to see once. Use whichever you find quicker, but know that both end at the same product because a factorisation into integers is unique up to order.
McDougal Littell Algebra 2 (Texas Edition), Ch. 4 Quadratic Functions and Factoring — Lesson 4.4 Solve ax^2 + bx + c = 0 by Factoring §4.4, p. 259 — Example 1, factored here by the ac-method
Picture it
Grouping can feel like symbol-pushing until you see where the two groups live. They are the two rows of the multiplication box you already know how to draw.
Figure (svg): The same multiplication box with its top row shaded blue and its bottom row shaded orange, a bracket beside each row reading 5x times x minus 3 and minus 2 times x minus 3, and a note that both rows end in the same bracket.
The top row is five x times the whole second bracket, and the bottom row is negative two times the same second bracket. Taking out a common factor from a row is just reading that row backwards.
So the bracket the two groups share is the column heading, and the factors you took out of the rows are the row headings. Grouping recovers the two headings from the four cells, which is exactly what factoring has to do.
That also tells you when grouping has gone wrong. If the two groups leave different brackets, the four cells you wrote down are not the cells of any single box, so the split of the middle term must have been wrong.
Draw the box the first few times you group. Once the correspondence is in your head you can group on a line, and you will still know what the line means.
Worked example
Here the ac-method earns its keep. The leading coefficient fifteen splits two ways and the constant eight splits three ways, so the arrangement table would run to a dozen rows.
\[ 15x^2 - 2x - 8 \]
Multiply the leading coefficient by the constant
Why: fifteen times negative eight is negative one hundred and twenty.
\[ ac = 15(-8) = -120 \]
Find two integers with product negative one hundred and twenty and sum negative two
Why: the product is negative, so the two disagree in sign, and they must be close together in size to add to only negative two.
\[ 10 \cdot (-12) = -120, \qquad 10 + (-12) = -2 \]
The search took one thought rather than twelve expansions. Two numbers about eleven apart whose product is one hundred and twenty: ten and twelve are the obvious candidates, and the sum decides which is negative.
Rewrite the middle term
Why: ten x minus twelve x is negative two x, so the value is unchanged.
\[ 15x^2 + 10x - 12x - 8 \]
Group and take out each pair's common factor
Why: five x comes out of the first pair and negative four out of the second.
\[ 5x(3x + 2) - 4(3x + 2) \]
Take out the shared bracket
Why: three x plus two is common to both terms.
\[ (3x + 2)(5x - 4) \]
Check by expanding
Why: fifteen x squared, then ten x and negative twelve x making negative two x, then two times negative four giving negative eight.
\[ (3x + 2)(5x - 4) = 15x^2 - 12x + 10x - 8 = 15x^2 - 2x - 8 \]
Notice the order of the two brackets does not matter, and neither does which pair you group first. What matters is that the two numbers you chose had the right product as well as the right sum.
McDougal Littell Algebra 2 (Texas Edition), Ch. 4 Quadratic Functions and Factoring — Lesson 4.4 Solve ax^2 + bx + c = 0 by Factoring §4.4, p. 263 — Exercise 8
Ranking
Five moves, and the order is not negotiable, because each one produces the thing the next one needs. Arrange them, then check your ordering against a trinomial you have already factored.
Put in order
Put the steps of the ac-method into the order you carry them out.
Why: The product of the leading coefficient and the constant has to exist before you can hunt for two numbers making it, and you cannot split the middle term until you know what to split it into. Grouping then needs four terms rather than three, and the shared bracket only appears once both pairs have had their common factor removed. Skipping the multiplication at the start is the usual failure: people look for two numbers multiplying to the constant alone, which is the rule for a leading coefficient of one.
The step people are most tempted to shorten is the hunt for the two integers. Finding two numbers with a given product and a given sum is the only genuinely searching move here, and it is worth writing the factor pairs of that product out when the numbers get large.
Notice which move carries the risk. Splitting the middle term changes how the expression looks without changing its value, so an error there is invisible until the two groups fail to leave the same bracket.
Use that as your checkpoint. If the two brackets after grouping do not match, go back to the pair of numbers and test both conditions on them again, rather than trying to force the brackets to agree.
Error analysis
Here is grouping applied with a split that satisfies one condition and not the other. The interesting thing is where the mistake becomes visible, which is two lines after it was made.
Annotate
On: \( 6x^2 + 7x - 3 = 6x^2 + 6x + x - 3 = 6x(x + 1) + (x - 3) \)
\[ 6x^2 + 9x - 2x - 3 = 3x(2x + 3) - (2x + 3) = (2x + 3)(3x - 1) \]
Check the repair by expanding: six x squared plus nine x minus two x minus three, which is six x squared plus seven x minus three. The original is recovered, so the corrected split was the right one.
The lesson is about which condition gets forgotten. The sum is the one you can see in the trinomial, so it is the one people check; the product is the one that makes the grouping work, so it is the one that matters.
Build the habit of stating both numbers with both conditions before you write anything: product negative eighteen, sum seven, so nine and negative two. Two seconds of bookkeeping against two wasted lines.
McDougal Littell Algebra 2 (Texas Edition), Ch. 4 Quadratic Functions and Factoring — Lesson 4.4 Solve ax^2 + bx + c = 0 by Factoring §4.4, p. 263 — the style of Exercise 31, applied to grouping
Section
Part 3
Concept
The book puts this in a single sentence: when factoring an expression, first check whether the terms have a common monomial factor. It reads like tidiness advice. It is actually the single biggest time saving in the lesson.
Figure (svg): On the left, two identical bars stacked to make a rectangle of height 2, each bar labelled three q squared minus seven q plus four. On the right, one of those bars written as a multiplication box for 3q minus 4 times q minus 1.
Read the left panel literally. Six q squared minus fourteen q plus eight is two copies of three q squared minus seven q plus four, stacked, so the height 2 is a genuine length and the picture is a real area for once.
Now count the work saved. Searching on a leading coefficient of six means the pairs six and one, and three and two, so two leading pairs. Searching on three means the single pair three and one. You have halved the table before starting it.
| what you search | leading pairs | arrangements of the constant | rows to test |
|---|---|---|---|
| 6q² − 14q + 8 as it stands | 6 and 1; 3 and 2 | four | eight |
| 3q² − 7q + 4 after the 2 comes out | 3 and 1 | three | three |
Eight rows against three, for the price of noticing that all three coefficients are even. That is the whole argument, and it is why the common factor sits at stage one of the procedure and not stage four.
McDougal Littell Algebra 2 (Texas Edition), Ch. 4 Quadratic Functions and Factoring — Lesson 4.4 Solve ax^2 + bx + c = 0 by Factoring §4.4, p. 260 — the Factoring Out Monomials note
Worked example
The book's own example of the saving. Work it in the order the book insists on, and notice that the search at the end is almost trivial.
\[ 6q^2 - 14q + 8 \]
Look for a common monomial factor
Why: all three coefficients are even and no power of q is shared by every term, so the common factor is 2.
\[ 6q^2 - 14q + 8 = 2(3q^2 - 7q + 4) \]
Say out loud what you have and have not done. The 2 is now outside and will stay there; the remaining job is to factor the trinomial inside the bracket, which has smaller numbers than the one you started with.
Factor the leading coefficient and the constant inside
Why: three is prime, four splits as one and four or as two and two, and both constants are negative because the constant is positive and the middle coefficient negative.
\[ a = 3 = 3 \cdot 1, \qquad c = 4 = (-1)(-4) = (-2)(-2) = (-4)(-1) \]
Test the arrangement with the four where the three can reach it
Why: three times four is twelve, which overshoots seven, so try the three against the one instead: three times one plus four is seven.
\[ (3q - 4)(q - 1) = 3q^2 - 3q - 4q + 4 = 3q^2 - 7q + 4 \]
Put the common factor back in front
Why: the 2 was never part of the trinomial being factored, so it simply rejoins the answer.
\[ 6q^2 - 14q + 8 = 2(3q - 4)(q - 1) \]
Never lose that 2. Dropping it gives an expression whose value is half of the one you were asked about, and it is the most common error in a question with a common factor.
Check by expanding everything
Why: the bracket product is three q squared minus seven q plus four, and doubling every term gives six q squared minus fourteen q plus eight.
\[ 2(3q^2 - 7q + 4) = 6q^2 - 14q + 8 \]
Compare with what a direct search would have cost. Without the 2 out first, the leading pairs six and one, and three and two, both had to be tried against four arrangements, which is eight expansions instead of two.
McDougal Littell Algebra 2 (Texas Edition), Ch. 4 Quadratic Functions and Factoring — Lesson 4.4 Solve ax^2 + bx + c = 0 by Factoring §4.4, p. 260 — Example 4(b)
Error analysis
The book's own error-analysis exercise, and it is worth more than it looks. The final line is beautifully tidy, and it is wrong by a factor that changes every value.
Annotate
On: \( 4x^2 - 36 = 4(x^2 - 36) = 4(x + 6)(x - 6) \)
\[ 4x^2 - 36 = 4(x^2 - 9) = 4(x + 3)(x - 3) \]
Every term, without exception. When you take a 4 out, write the bracket by dividing each coefficient by four, and if one of them does not divide exactly then 4 was not a common factor in the first place.
The numerical test is the habit to steal from this slide. Substituting one convenient value into both your answer and the original takes five seconds and catches exactly this class of error, which expanding also catches but more slowly.
Notice that the error would have been impossible had the person expanded. Four times x plus six, times x minus six, is four x squared minus one hundred and forty-four, and nobody writes that down and fails to notice.
McDougal Littell Algebra 2 (Texas Edition), Ch. 4 Quadratic Functions and Factoring — Lesson 4.4 Solve ax^2 + bx + c = 0 by Factoring §4.4, p. 263 — Exercise 31
Worked example
A negative leading coefficient is where signs get lost, so there is a standing recommendation: take the minus sign out with the common factor, so that the trinomial you search is the friendly way up.
\[ -36n^2 + 48n - 15 \]
Find the common monomial factor, and take the sign with it
Why: thirty-six, forty-eight and fifteen are all divisible by three, and the leading coefficient is negative, so take out negative three.
\[ -36n^2 + 48n - 15 = -3(12n^2 - 16n + 5) \]
Check that bracket term by term before going on, because this is where a sign goes missing. Negative three times twelve n squared is negative thirty-six n squared; negative three times negative sixteen n is plus forty-eight n; negative three times five is negative fifteen. All three match.
Factor the leading coefficient and the constant inside
Why: twelve splits as twelve and one, six and two, or four and three; five is prime, and both constants are negative.
\[ a = 12, \qquad c = 5 = (-1)(-5) = (-5)(-1) \]
Look for the arrangement that reaches negative sixteen
Why: six times one plus two times five is sixteen, so try six and two as the leading pair with negative five beside the two.
\[ (6n - 5)(2n - 1) = 12n^2 - 6n - 10n + 5 = 12n^2 - 16n + 5 \]
Write the whole factorisation
Why: the negative three stays outside, in front of the two brackets.
\[ -36n^2 + 48n - 15 = -3(6n - 5)(2n - 1) \]
Check by substituting a value
Why: at n equal to one the original is negative thirty-six plus forty-eight minus fifteen, which is negative three; the factored form gives negative three times one times one, which is also negative three.
\[ n = 1: \quad -36 + 48 - 15 = -3, \qquad -3(1)(1) = -3 \]
Substituting one is the cheapest check there is, because every bracket becomes a small number. It will not catch every error, but it catches a sign error almost every time.
McDougal Littell Algebra 2 (Texas Edition), Ch. 4 Quadratic Functions and Factoring — Lesson 4.4 Solve ax^2 + bx + c = 0 by Factoring §4.4, p. 263 — Exercise 29
Trap
Factoring negative sixteen n squared plus twelve n, a line you will see often:
\[ -16n^2 + 12n = 4n(-4n + 3) \]
This is actually correct, and it is still a bad habit, because the next person to factor that bracket has a negative leading coefficient to deal with all over again.
The trouble comes when the same instinct meets a trinomial and the minus sign gets left behind on only some of the terms.
Take the sign out with the factor instead.
\[ -16n^2 + 12n = -4n(4n - 3) \]
Now every remaining coefficient is positive, and any further factoring is the familiar kind.
Check it the same way as always: negative four n times four n is negative sixteen n squared, and negative four n times negative three is plus twelve n.
The rule to write on your page is short. When the leading coefficient is negative, make the common factor negative too, then check the bracket term by term before you use it.
Fill the middle
Large coefficients, and they are large for a reason: the exercise is built so that not taking the common factor out is genuinely painful. Look at all three numbers before you touch anything.
Fill in the blanks
20x^2 + 124x + 24 = 4(5x + 1)(x + 6)
Why: Twenty, one hundred twenty-four and twenty-four are all divisible by four, leaving five x squared plus thirty-one x plus six. Five is prime and six splits as one and six or as two and three, and five times six plus one is thirty-one, so the arrangement pairs the five with the six. Expanding back: five x squared plus thirty x plus x plus six, times four, is twenty x squared plus one hundred twenty-four x plus twenty-four.
Imagine the alternative. Twenty splits as twenty and one, ten and two, or five and four, and twenty-four splits six ways, which is a table of nearly forty rows. Nobody finishes that, which is exactly the experience people call endless trial and error.
The common factor is what turns it into a table of six rows, and the size estimate then picks the row on the first try, because five times six is thirty and you need thirty-one.
Check the value of the answer at a convenient number. At x equal to one the original is twenty plus one hundred twenty-four plus twenty-four, which is one hundred sixty-eight, and the factored form gives four times six times seven, which is also one hundred sixty-eight.
McDougal Littell Algebra 2 (Texas Edition), Ch. 4 Quadratic Functions and Factoring — Lesson 4.4 Solve ax^2 + bx + c = 0 by Factoring §4.4, p. 263 — Exercise 27
Sorting
Six expressions from the exercise set. For each one, decide what the greatest common monomial factor is before any other thought, and note that sometimes the answer is nothing at all.
Sort into buckets
Sort each expression by what its greatest common monomial factor looks like.
The two with no common factor are worth noticing. A trinomial can have a leading coefficient of nine or of five and still have nothing to take out, so stage one of the procedure is a check, not a guarantee.
The two with a variable coming out are a different shape of problem altogether. With no constant term the expression is already a product, and after the monomial comes out only a linear bracket is left, so there is no searching to do at all.
Take the greatest common factor, not just any common factor. Taking a 2 out of twelve x squared minus four x minus forty leaves six x squared minus two x minus twenty, which still has a 2 in it, and you will only have to do the job twice.
McDougal Littell Algebra 2 (Texas Edition), Ch. 4 Quadratic Functions and Factoring — Lesson 4.4 Solve ax^2 + bx + c = 0 by Factoring §4.4, p. 263 — Exercises 22 to 29
Worked example
The chapter's challenge exercises look like a new topic and are not. Once the common factor is out, what remains is the trinomial you have been factoring all along.
\[ 2x^3 - 5x^2 + 3x \]
Find the common monomial factor
Why: the coefficients two, five and three share nothing, but every term carries at least one x, so x comes out.
\[ 2x^3 - 5x^2 + 3x = x(2x^2 - 5x + 3) \]
That single move is what makes the problem ordinary. A cubic with three terms is not something you have a method for; a monomial times a quadratic trinomial is.
Factor the trinomial inside
Why: two is prime, three is prime, and the constant is positive with a negative middle coefficient, so both constants are negative.
\[ a = 2 = 2 \cdot 1, \qquad c = 3 = (-1)(-3) = (-3)(-1) \]
Test the arrangement that reaches negative five
Why: two times three plus one is seven, which is too big, so pair the two with the one instead: two times one plus three is five.
\[ (2x - 3)(x - 1) = 2x^2 - 2x - 3x + 3 = 2x^2 - 5x + 3 \]
Write the complete factorisation
Why: three factors now, and the x in front is one of them.
\[ 2x^3 - 5x^2 + 3x = x(2x - 3)(x - 1) \]
If this were an equation set equal to zero, that leading x would be a third solution. Dividing it away rather than factoring it out is how people lose the root at zero, which Part 5 comes back to.
Check by substituting a value
Why: at x equal to two the original is sixteen minus twenty plus six, which is two; the factored form gives two times one times one, which is also two.
\[ x = 2: \quad 16 - 20 + 6 = 2, \qquad 2(1)(1) = 2 \]
Choose your test value to avoid zero. At x equal to zero every side is zero whatever mistake you have made, so it tells you nothing at all.
McDougal Littell Algebra 2 (Texas Edition), Ch. 4 Quadratic Functions and Factoring — Lesson 4.4 Solve ax^2 + bx + c = 0 by Factoring §4.4, p. 264 — Exercise 59
Comparison
One last look at the arithmetic of the saving, this time as a table you complete yourself. Count the rows each route would ask you to expand.
Comparison matrix
| route | leading pairs to try | arrangements of the constant | rows to expand |
|---|---|---|---|
| factor 6q² − 14q + 8 straight away | 6 and 1; 3 and 2 | four | 8 |
| take the 2 out, then factor 3q² − 7q + 4 | 3 and 1 | three | 3 |
Multiply the two middle columns in each row and you have the number of expansions. Two leading pairs against four arrangements is eight; one leading pair against three arrangements is three.
Three is a table you write and finish. Eight is a table most people abandon halfway through and start guessing instead, which is where the feeling of endlessness comes from.
The saving compounds with the size of the common factor. Take a four out and the leading coefficient drops by a factor of four, which usually removes at least one leading pair and often two.
There is a second benefit that does not show in the table. Smaller numbers make the size estimate sharper, so you pick the right row sooner even within the shorter list.
Section
Part 4
Concept
Lesson 4.3 gave you two patterns for a leading coefficient of one. They still hold when the leading coefficient is a square, because the pattern never cared whether the thing being squared was a bare x.
\[ u^2 - v^2 = (u + v)(u - v) \]
Figure (svg): On the left a square of side 3x with an eight by eight corner cut out of it, the remaining L-shape shaded. On the right that same L-shape re-laid as a rectangle whose sides are 3x plus 8 and 3x minus 8.
The picture is the proof. A square of side three x with an eight by eight corner removed has area nine x squared minus sixty-four; slide the removed strip round and the same pieces form a rectangle of sides three x plus eight and three x minus eight.
So whenever a and the constant are both perfect squares and the sign between them is a minus, you can write the answer down with no search at all. Nine x squared is three x all squared, sixty-four is eight squared, and the factorisation follows.
Check first that the middle term really is absent. Nine x squared minus sixty-four has no x term, which is what lets the two middle products cancel, and a difference of squares always has that shape.
McDougal Littell Algebra 2 (Texas Edition), Ch. 4 Quadratic Functions and Factoring — Lesson 4.4 Solve ax^2 + bx + c = 0 by Factoring §4.4, p. 260 — the Factoring Special Products note
Notation
The second pattern is the one that gets misapplied, so take its statement apart before using it. Every symbol in it is doing work, and the middle term is where the test lives.
Annotate
On: \( u^2 + 2uv + v^2 = (u + v)^2 \)
Exactly is the word to underline. If the middle coefficient were nineteen or twenty-one instead of twenty, the trinomial would not be a perfect square, and no amount of adjusting the brackets would make it one.
Work in the direction the notes are written when you are checking. Find u from the leading term, find v from the constant, compute twice u v, and compare it with the middle term you actually have.
The minus version catches people because the constant is still positive. Thirty-six w squared minus twelve w plus one is a perfect square, with v equal to one and a minus sign inside the bracket, and the plus one at the end is not a contradiction.
McDougal Littell Algebra 2 (Texas Edition), Ch. 4 Quadratic Functions and Factoring — Lesson 4.4 Solve ax^2 + bx + c = 0 by Factoring §4.4, p. 260 — Example 3 and the Factoring Special Products note
Worked example
The book's third example is three parts, each one a pattern recognised rather than searched for. Work them in order and say out loud which pattern you are using before writing the answer.
\[ \text{(a) } 9x^2 - 64 \qquad \text{(b) } 4y^2 + 20y + 25 \qquad \text{(c) } 36w^2 - 12w + 1 \]
Part (a): write each term as a square
Why: nine x squared is three x all squared, and sixty-four is eight squared, with a minus sign between them.
\[ 9x^2 - 64 = (3x)^2 - 8^2 \]
Apply the difference of two squares
Why: the same expression appears in both brackets, once with a plus and once with a minus.
\[ 9x^2 - 64 = (3x + 8)(3x - 8) \]
Part (b): identify u and v, then test the middle term
Why: u is two y and v is five, and twice their product is twenty y, which matches the middle term exactly.
\[ 4y^2 + 20y + 25 = (2y)^2 + 2(2y)(5) + 5^2 = (2y + 5)^2 \]
That middle line is worth writing every time, even when you are confident. It is the line that proves the pattern applies, and without it you are asserting the answer rather than deriving it.
Part (c): the same test with a minus sign
Why: u is six w and v is one, and twice their product is twelve w, which is the size of the middle term; the middle term is negative, so the bracket carries a minus.
\[ 36w^2 - 12w + 1 = (6w)^2 - 2(6w)(1) + 1^2 = (6w - 1)^2 \]
Check all three by expanding
Why: the first gives nine x squared minus twenty-four x plus twenty-four x minus sixty-four; the second and third each square their bracket and reproduce the trinomial.
\[ (2y + 5)^2 = 4y^2 + 10y + 10y + 25, \qquad (6w - 1)^2 = 36w^2 - 6w - 6w + 1 \]
Look at what the expansion of a squared bracket does: the two middle products are identical, which is exactly why the middle coefficient is twice a product rather than a sum of two different things.
McDougal Littell Algebra 2 (Texas Edition), Ch. 4 Quadratic Functions and Factoring — Lesson 4.4 Solve ax^2 + bx + c = 0 by Factoring §4.4, p. 260 — Example 3
Picture it
The perfect square pattern deserves its own picture, because the reason the middle coefficient is doubled is visible in it and nowhere else.
Figure (svg): A square of side 2y plus 5, divided into four pieces: a 2y by 2y square labelled 4y squared, two strips each labelled 10y, and a five by five square labelled 25.
The whole figure is one square of side two y plus five. Splitting each side at the same place cuts it into four pieces: a two y by two y square, a five by five square, and two strips that are each two y by five.
Those two strips are the middle term. They have the same area because the square is symmetric about its diagonal, so the middle term is two lots of ten y, which is twenty y.
That is why the test is twice the product and not simply the product. There are always two strips in a square, so the middle term is always doubled, and a trinomial with only one strip's worth cannot be a square.
This picture is a genuine area, unlike the boxes in Part 1, because every length in it is positive. When the middle term is negative the algebra still works, but you are back to bookkeeping rather than geometry.
Prediction
This is the trap the pattern sets. Both outer terms are perfect squares, four x squared and twenty-five, and it is tempting to conclude that the whole trinomial must be a perfect square trinomial.
Figure (svg): The same square, now for four x squared plus twenty-one x plus twenty-five, with a thin red strip standing outside the square to show the extra x that cannot be placed.
Predict first
Four x squared plus twenty-one x plus twenty-five has perfect squares at both ends. Is it a perfect square trinomial?
Correct: No, the middle term would have to be 20x
Why: With u equal to two x and v equal to five, twice the product is twenty x. The trinomial has twenty-one x, which is one x too many, so no bracket squared can produce it. The picture shows the extra strip standing outside the completed square with nowhere to go, and in fact this trinomial does not factor over the integers at all.
The picture makes the failure concrete. Twenty x is exactly the two strips the square needs, and the twenty-first x is a sliver that no rearrangement of a square can absorb.
So the outer terms being squares is a permission to test the pattern, never a conclusion that it holds. Compute twice the product, compare, and only then commit.
It is worth knowing what would have happened with a nineteen or a twenty-one. Neither gives a perfect square, and neither factors at all here, which is a reminder that most trinomials do not factor over the integers. The ones in a textbook exercise are chosen.
McDougal Littell Algebra 2 (Texas Edition), Ch. 4 Quadratic Functions and Factoring — Lesson 4.4 Solve ax^2 + bx + c = 0 by Factoring §4.4, p. 263 — Exercise 2, on how a and c help you decide
Fill the middle
A perfect square with a minus sign and with both ends larger than usual. Find u from the leading term and v from the constant, then test the middle term before filling anything in.
Fill in the blanks
25s^2 - 80s + 64 = (5s - 8)^2
Why: Twenty-five s squared is five s all squared and sixty-four is eight squared, so u is five s and v is eight. Twice their product is twice five s times eight, which is eighty s, and that is exactly the middle term, so the pattern applies. The middle term is negative, so the bracket carries a minus sign. Expanding checks it: twenty-five s squared minus forty s minus forty s plus sixty-four.
The test is the whole slide. Had the middle coefficient been seventy-nine or eighty-one, this expression would not be a perfect square, and you would have gone back to the four-integer search with a leading coefficient of twenty-five.
Do not let the size of the numbers put you off. Eighty looks large until you notice it is twice forty, and forty is five times eight, which are the two numbers you already had.
Notice that the constant stays positive with a minus sign inside the bracket. Negative eight squared is positive sixty-four, so the sign of the constant tells you nothing about the sign inside; only the middle term does.
McDougal Littell Algebra 2 (Texas Edition), Ch. 4 Quadratic Functions and Factoring — Lesson 4.4 Solve ax^2 + bx + c = 0 by Factoring §4.4, p. 260 — Guided Practice 10
Two truths and a lie
Sort these four statements into true and false. Each one is a sentence somebody has said in a lesson, and three of them are worth remembering.
Sort into buckets
Sort each claim by whether it is true.
The claim about a sum of two squares is worth dwelling on. A difference of squares factors and a sum does not, which is the sharpest reminder available that the sign between the terms is part of the pattern.
Test that claim yourself rather than accepting it. Factoring x squared plus sixty-four over the integers would need two numbers multiplying to sixty-four and adding to zero, and two numbers adding to zero are opposites, whose product is negative.
The false claim is the one you will be tempted by under time pressure, because it turns a test into a shortcut. Spending the one line that computes twice the product is what protects you.
McDougal Littell Algebra 2 (Texas Edition), Ch. 4 Quadratic Functions and Factoring — Lesson 4.4 Solve ax^2 + bx + c = 0 by Factoring §4.4, p. 260 — Example 3 and the Guided Practice that follows it
Worked example
This one joins Part 3 to Part 4, and it is the reason the procedure puts the common factor before the patterns. Look at it once with the factor out and once without.
\[ 36n^2 - 9 \]
Take out the common monomial factor
Why: thirty-six and nine are both divisible by nine, and no power of n is shared, so nine comes out.
\[ 36n^2 - 9 = 9(4n^2 - 1) \]
Recognise the bracket as a difference of two squares
Why: four n squared is two n all squared and one is one squared, with a minus between them.
\[ 9(4n^2 - 1) = 9\bigl((2n)^2 - 1^2\bigr) \]
Apply the pattern
Why: the same expression twice, once with a plus and once with a minus.
\[ 36n^2 - 9 = 9(2n + 1)(2n - 1) \]
Now look at what happens if you skip the common factor. Thirty-six n squared is six n all squared and nine is three squared, so the pattern still applies directly, and you get six n plus three, times six n minus three.
Compare the two routes
Why: the second route gives brackets that still contain a common factor of three each, and taking both out recovers the nine.
\[ (6n + 3)(6n - 3) = 3(2n + 1) \cdot 3(2n - 1) = 9(2n + 1)(2n - 1) \]
Check by substituting a value
Why: at n equal to one the original is thirty-six minus nine, which is twenty-seven, and the factored form gives nine times three times one, which is also twenty-seven.
\[ n = 1: \quad 36 - 9 = 27, \qquad 9(3)(1) = 27 \]
Both routes reach the same place, but only one of them is finished when it stops. Factored completely means no bracket still has a common factor inside it, which is worth checking as a last step every time.
McDougal Littell Algebra 2 (Texas Edition), Ch. 4 Quadratic Functions and Factoring — Lesson 4.4 Solve ax^2 + bx + c = 0 by Factoring §4.4, p. 260 — Guided Practice 12
Section
Part 5
Concept
Everything so far has been factoring. Solving adds one idea, and you already have it: if a product of two things is zero, one of them is zero. That is the zero product property from Lesson 4.3, and nothing about it changes when the leading coefficient grows.
\[ 3x^2 + 10x - 8 = 0 \;\Longrightarrow\; (3x - 2)(x + 4) = 0 \;\Longrightarrow\; 3x - 2 = 0 \;\text{ or }\; x + 4 = 0 \]
What does change is the amount of setting up. An equation rarely arrives with a zero on the right and no common factor, so there are two preparatory moves to make before factoring is even possible.
The second move deserves a word about why it is safe here and not always. Dividing an equation by a nonzero number changes nothing about its solutions, because the right-hand side is zero and zero divided by five is still zero. Dividing by something containing a variable is a different matter, and it is the trap two slides from the end of this part.
McDougal Littell Algebra 2 (Texas Edition), Ch. 4 Quadratic Functions and Factoring — Lesson 4.4 Solve ax^2 + bx + c = 0 by Factoring §4.4, p. 261 — the Solving Quadratic Equations note
Worked example
An equation already in standard form, with nothing to divide out, so the only work is the factoring you have been practising and then one line per bracket.
\[ 3x^2 + 10x - 8 = 0 \]
Factor the left-hand side
Why: three is prime and eight splits as one and eight or two and four; the constant is negative so the two numbers disagree in sign, and three times four minus two is ten.
\[ (3x - 2)(x + 4) = 0 \]
Apply the zero product property
Why: a product is zero only when one of its factors is zero, so write one equation per bracket.
\[ 3x - 2 = 0 \quad \text{or} \quad x + 4 = 0 \]
Two brackets, two equations, two answers. Notice that each equation is linear, which is the whole point of factoring: a quadratic you cannot solve directly has become two you can.
Solve each linear equation
Why: add two and divide by three in the first; subtract four in the second.
\[ x = \dfrac{2}{3} \quad \text{or} \quad x = -4 \]
Check both answers in the original equation
Why: three times four ninths is four thirds, plus twenty thirds, is twenty-four thirds, which is eight, and eight minus eight is zero; and three times sixteen is forty-eight, minus forty, minus eight, is zero.
\[ 3\left(\tfrac23\right)^2 + 10\left(\tfrac23\right) - 8 = \tfrac43 + \tfrac{20}{3} - 8 = 0 \]
\[ 3(-4)^2 + 10(-4) - 8 = 48 - 40 - 8 = 0 \]
Substitute both, always. A bracket that was factored slightly wrongly often still gives one correct root, so checking only one answer is barely a check at all.
McDougal Littell Algebra 2 (Texas Edition), Ch. 4 Quadratic Functions and Factoring — Lesson 4.4 Solve ax^2 + bx + c = 0 by Factoring §4.4, p. 261 — Example 5(a)
Picture it
An equation set equal to zero is a question about where a graph meets the horizontal axis. Seeing that once makes the zero product property feel inevitable rather than clever.
Figure (svg): The parabola y equals three x squared plus ten x minus eight, crossing the horizontal axis at minus four and at two thirds, both crossings marked.
The curve is the function three x squared plus ten x minus eight, and the two marked points are exactly the two answers from the last slide. Solving the equation is asking the curve where its height is zero.
Each bracket contributes one crossing. Three x minus two is zero at two thirds, and x plus four is zero at negative four, and the product of the two is zero at each of those places and nowhere else.
The picture also tells you how many answers to expect. An upward parabola that dips below the axis has to cross it twice, so two answers is the normal case, and it is the case a successful factorisation always describes.
Keep the graph in mind when an answer looks odd. A root outside the range you expected, or a positive root where the picture crosses on the negative side, is caught by a glance at the curve long before it is caught by algebra.
Step zero
Before solving anything, decide the opening move. This equation is not in standard form, and choosing the wrong first step is where most of the marks go on questions like this one.
\[ 5p^2 - 16p + 15 = 4p - 5 \]
Discussion prompt
Name the first move you would make, and say why factoring the left-hand side straight away would be useless.
This is the single most valuable habit in Part 5. The zero product property is a statement about zero, so an equation that is not equal to zero has to be moved before the property is available.
There is a bonus once you have collected. All three coefficients of the collected equation are divisible by five, which means the next move shrinks the numbers, and that is the second stage of the procedure arriving on cue.
Write the collection out rather than doing it in your head. Two terms move and both change sign, and a sign lost at this stage produces a completely different equation that will still factor neatly and give you two wrong answers.
McDougal Littell Algebra 2 (Texas Edition), Ch. 4 Quadratic Functions and Factoring — Lesson 4.4 Solve ax^2 + bx + c = 0 by Factoring §4.4, p. 261 — Example 5(b), step 1
Worked example
The equation from the previous slide, carried through. It has three lessons in it: reach standard form, divide out, and then notice what a squared bracket means.
\[ 5p^2 - 16p + 15 = 4p - 5 \]
Write the equation in standard form
Why: subtract four p and add five on both sides, so that the right-hand side becomes zero.
\[ 5p^2 - 20p + 20 = 0 \]
Divide each side by the common factor
Why: five divides all three coefficients, and dividing an equation whose right side is zero by a nonzero number changes none of its solutions.
\[ p^2 - 4p + 4 = 0 \]
Look at what that division bought. A leading coefficient of five became one, which means the search is back to Lesson 4.3 territory, and in this case it is a pattern you can recognise on sight.
Factor the left-hand side
Why: one and four are both perfect squares and twice one times two is four, so this is a perfect square trinomial.
\[ (p - 2)^2 = 0 \]
Apply the zero product property
Why: the two factors are the same bracket, so both equations are the same equation, and there is only one solution.
\[ p - 2 = 0 \;\Longrightarrow\; p = 2 \]
A repeated root is not half an answer. It is one value of p that makes both factors zero at once, and that is why the book's note says that if the square of an expression is zero then the expression itself must be zero.
Check in the original equation, not the rearranged one
Why: the left side is twenty minus thirty-two plus fifteen, which is three, and the right side is eight minus five, which is also three.
\[ 5(2)^2 - 16(2) + 15 = 3, \qquad 4(2) - 5 = 3 \]
Checking in the original is what verifies the collecting as well as the factoring. Had a sign gone missing when you moved four p across, the rearranged equation would still check out perfectly.
McDougal Littell Algebra 2 (Texas Edition), Ch. 4 Quadratic Functions and Factoring — Lesson 4.4 Solve ax^2 + bx + c = 0 by Factoring §4.4, p. 261 — Example 5(b)
Picture it
Two brackets that are the same bracket sounds like a technicality. On a graph it is a clear and memorable picture.
Figure (svg): Two parabolas. The solid one, p squared minus four p plus four, touches the horizontal axis once at p equals two; the dashed one, p squared minus four p plus three, crosses it twice, at one and at three.
The solid curve is the bracket p minus two, squared. It comes down to the axis at p equal to two, touches it, and goes back up without ever crossing, so there is exactly one place where the height is zero.
The dashed curve is the same parabola pushed down by one. Its two crossings sit either side of the touching point, and they have pulled apart. Push it down further and they separate more; lift it back and they merge again into one.
So a repeated root is the boundary case between two answers and none. That idea is the whole content of the discriminant in Lesson 4.8, and it is worth meeting it as a picture first.
On the practical side, when a factorisation comes out as a bracket squared, write down the one answer and say that it is repeated. Writing it twice is not wrong, but saying it is repeated shows you know why.
Worked example
An equation that is already in standard form and where the second stage of the procedure does real work. Look at the three coefficients before you factor anything.
\[ 6x^2 - 3x - 63 = 0 \]
Divide each side by the common factor
Why: six, three and sixty-three are all divisible by three, and the right-hand side is zero, so dividing is safe.
\[ 2x^2 - x - 21 = 0 \]
That single division dropped the leading coefficient from six to two, which removes the leading pair three and two from the search and leaves only two and one. The numbers on the constant shrank as well.
Factor the left-hand side
Why: two is prime; twenty-one splits as one and twenty-one or three and seven; the constant is negative so the signs disagree, and two times three minus seven is negative one.
\[ (2x - 7)(x + 3) = 0 \]
Apply the zero product property
Why: one equation per bracket.
\[ 2x - 7 = 0 \quad \text{or} \quad x + 3 = 0 \]
Solve each one
Why: add seven and halve in the first; subtract three in the second.
\[ x = \dfrac{7}{2} \quad \text{or} \quad x = -3 \]
Check both in the original equation
Why: six times forty-nine quarters is seventy-three and a half, minus ten and a half, minus sixty-three, is zero; and six times nine is fifty-four, plus nine, minus sixty-three, is zero.
\[ 6\left(\tfrac72\right)^2 - 3\left(\tfrac72\right) - 63 = \tfrac{147}{2} - \tfrac{21}{2} - 63 = 0 \]
\[ 6(-3)^2 - 3(-3) - 63 = 54 + 9 - 63 = 0 \]
Both check, and notice that the fractional root is the one worth checking carefully, since a fraction squared is where arithmetic slips happen.
McDougal Littell Algebra 2 (Texas Edition), Ch. 4 Quadratic Functions and Factoring — Lesson 4.4 Solve ax^2 + bx + c = 0 by Factoring §4.4, p. 262 — Guided Practice 19
Fill the middle
Both preparatory moves are needed here: collect to standard form, then divide out. Do them in that order and the factoring at the end is easy.
Fill in the blanks
12x^2 + 6x - 6 = 0 \;\Longrightarrow\; 6(2x - 1)(x + 1) = 0 \;\Longrightarrow\; x = 1/2 \text___ x = -1
Why: Subtracting x and eight from both sides gives twelve x squared plus six x minus six equals zero, and all three coefficients are divisible by six, leaving two x squared plus x minus one. That factors as two x minus one times x plus one, so the answers are one half and negative one. Check the first in the original: twelve times a quarter plus three and a half plus two is eight and a half, and x plus eight is also eight and a half.
The six is what makes the rest painless. Without it you would be searching a leading coefficient of twelve, which splits four ways, against a constant of six, which splits three ways, for a table of about twenty rows.
Notice that dividing by six does not change the answers. The equation two x squared plus x minus one equals zero has exactly the same solutions as the one before it, because zero divided by six is still zero.
Check the second answer as well. At negative one the left side of the original is twelve minus seven plus two, which is seven, and the right side is negative one plus eight, which is also seven.
McDougal Littell Algebra 2 (Texas Edition), Ch. 4 Quadratic Functions and Factoring — Lesson 4.4 Solve ax^2 + bx + c = 0 by Factoring §4.4, p. 262 — Guided Practice 20
Worked example
A short equation with a genuine hazard in it. The temptation is to divide both sides by x, and doing so quietly throws away one of the two answers.
\[ x = 4x^2 - 15x \]
Write the equation in standard form
Why: bring the lone x across so the right-hand side is zero; collecting the two x terms gives negative sixteen x.
\[ 0 = 4x^2 - 16x \]
Take out the common monomial factor
Why: four and sixteen share a four, and both terms carry an x, so four x comes out.
\[ 0 = 4x(x - 4) \]
Keep the four x as a factor rather than dividing it away. It is one of the two brackets, and setting it to zero is what produces the answer at zero.
Apply the zero product property
Why: the product is zero when the monomial is zero or when the bracket is zero.
\[ 4x = 0 \quad \text{or} \quad x - 4 = 0 \]
Solve each
Why: the first gives zero, the second gives four.
\[ x = 0 \quad \text{or} \quad x = 4 \]
Check both in the original equation
Why: at zero both sides are zero; at four the right side is sixty-four minus sixty, which is four, and the left side is four.
\[ x = 4: \quad 4(4)^2 - 15(4) = 64 - 60 = 4 \]
Notice how easy the answer at zero is to miss and how easy it is to check. Whenever an equation has no constant term, zero is a solution, and it should appear in your answer.
McDougal Littell Algebra 2 (Texas Edition), Ch. 4 Quadratic Functions and Factoring — Lesson 4.4 Solve ax^2 + bx + c = 0 by Factoring §4.4, p. 264 — Exercise 57
Trap
Solving the same equation, a tidier-looking route:
\[ 4x^2 - 16x = 0 \]
\[ 4x - 16 = 0 \]
Dividing every term by x looks exactly like the legal division by a number two slides ago, and the answer it gives, four, is genuinely one of the answers.
But one answer has vanished, and nothing in the working says so.
Factor instead of dividing, and both answers survive.
\[ 4x(x - 4) = 0 \]
Dividing by x is only legal when x is not zero, and zero was one of the solutions, so the division assumed away the very thing you were looking for.
\[ x = 0 \;\text{ or }\; x = 4 \]
The rule to keep: never divide an equation by something that could be zero. Take it out as a factor, and it will tell you its own root.
Worked example
The first problem where you have to write the equation before you can solve it. Read the situation, draw it, and only then reach for algebra.
A finished quilt measures five feet by four feet, and there are ten square feet of fabric left over for a decorative border of uniform width all the way round. How wide should the border be?
Figure (svg): A five foot by four foot quilt drawn to scale with a border of uniform width x all the way round it, the border shaded and the outer dimensions marked five plus two x and four plus two x.
Name the unknown and write the outer dimensions
Why: let x be the border's width in feet; the border adds x on each of two opposite sides, so each overall dimension grows by two x.
\[ \text{outer: } (5 + 2x) \text{ by } (4 + 2x) \]
Write a verbal model as an equation
Why: the area of the border is the whole area less the area of the quilt itself, and that has to equal the ten square feet available.
\[ 10 = (5 + 2x)(4 + 2x) - (5)(4) \]
Expand the product
Why: five times four, five times two x, two x times four, and two x times two x.
\[ 10 = 20 + 10x + 8x + 4x^2 - 20 \]
Write the equation in standard form
Why: the twenty and the negative twenty cancel, and the ten moves across.
\[ 0 = 4x^2 + 18x - 10 \]
Divide each side by the common factor
Why: all three coefficients are even, and the right-hand side is zero.
\[ 0 = 2x^2 + 9x - 5 \]
Factor and split
Why: two is prime and five is prime with a negative constant, so the signs disagree, and two times five minus one is nine.
\[ 0 = (2x - 1)(x + 5) \;\Longrightarrow\; x = \tfrac12 \;\text{ or }\; x = -5 \]
Two mathematical answers, one physical answer. A border cannot have a width of negative five feet, so that root is rejected, and the width is one half of a foot, which is six inches.
Check by computing the border area directly
Why: with a border of half a foot the outer rectangle is six feet by five feet, an area of thirty square feet, and the quilt itself is twenty, so the border uses ten square feet, exactly the fabric available.
\[ (6)(5) - (5)(4) = 30 - 20 = 10 \]
Check in the situation and not only in the equation. An arithmetic slip while expanding would give an equation that solves perfectly and answers the wrong question, and only the direct area calculation catches that.
McDougal Littell Algebra 2 (Texas Edition), Ch. 4 Quadratic Functions and Factoring — Lesson 4.4 Solve ax^2 + bx + c = 0 by Factoring §4.4, p. 261 — Example 6
Real world
The same shape of problem with one extra condition, and it is the condition that keeps the algebra to a single unknown. Read it carefully before writing anything.
A twenty-five inch by twenty-one inch painting is to be matted. The mat is to be twice as wide at the left and right as it is at the top and bottom, and there are seven hundred and fourteen square inches of mat available.
Figure (svg): A twenty-five inch by twenty-one inch painting with a mat round it, the mat twice as wide at the left and right, of width two x, as it is at the top and bottom, of width x.
Discussion prompt
Let x be the width of the mat at the top and bottom. Write the outer dimensions, form an equation, solve it, and state both widths.
Notice the choice made at the start. Calling the smaller width x keeps the side width as two x rather than as x over two, and whole-number coefficients are easier to factor than fractional ones.
Notice also which dimension gains four x. The sides are the left and right, so they add to the twenty-five inch measurement, and there are two of them at two x each.
The check at the end is a genuine area calculation rather than a substitution into the quadratic, and that is deliberate. It tests the modelling as well as the algebra.
McDougal Littell Algebra 2 (Texas Edition), Ch. 4 Quadratic Functions and Factoring — Lesson 4.4 Solve ax^2 + bx + c = 0 by Factoring §4.4, p. 264 — Exercise 66
Section
Part 6
Concept
Factoring has one more use, and it is the one that pays for itself in word problems. Once a quadratic is factored you can read off its maximum or minimum without completing a square or differentiating anything.
\[ y = a(x - p)(x - q) \qquad \text{intercept form} \]
Figure (svg): A downward parabola crossing the horizontal axis at p and at q, with a dashed vertical line through its peak and two equal arrows showing that the peak is the same distance from each crossing.
The two zeros are p and q, because those are the values that make one bracket vanish. A parabola is symmetric, so reflecting it in its own axis of symmetry must swap the two zeros, which forces that axis to run exactly midway between them.
\[ x = \dfrac{p + q}{2} \]
And the vertex sits on the axis of symmetry, so the maximum or minimum happens at the average of the zeros. Substitute that value back into the function and you have the extreme value itself.
So the recipe is short: factor into intercept form, average the two zeros, and substitute. Three lines, and no new machinery beyond the factoring you have spent this lesson on.
McDougal Littell Algebra 2 (Texas Edition), Ch. 4 Quadratic Functions and Factoring — Lesson 4.4 Solve ax^2 + bx + c = 0 by Factoring §4.4, p. 262 — the Factoring and Zeros note
Worked example
The book's revenue problem, and the reason the lesson exists in the form it does. A monthly teen magazine has twenty-eight thousand subscribers when it charges ten dollars a year, and each one dollar increase loses about two thousand subscribers.
Define the variables
Why: let x be the price increase in dollars, and let R of x be the annual revenue in dollars.
\[ x = \text{price increase}, \qquad R(x) = \text{annual revenue} \]
Write a verbal model as a function
Why: revenue is the number of subscribers multiplied by the price each one pays, and both of those depend on x.
\[ R(x) = (28\,000 - 2000x)(10 + x) \]
Check each bracket against the story before going on. At an increase of zero the first bracket is twenty-eight thousand subscribers and the second is ten dollars, which is the situation described, so the model starts in the right place.
Factor into intercept form
Why: take negative two thousand out of the first bracket, which turns twenty-eight thousand minus two thousand x into negative two thousand times x minus fourteen.
\[ R(x) = -2000(x - 14)(x + 10) \]
Read off the zeros and average them
Why: the brackets vanish at fourteen and at negative ten, and the vertex lies midway between them.
\[ x = \dfrac{14 + (-10)}{2} = 2 \]
Negative ten is not a price increase anybody would choose, and it is still needed. The average uses both zeros, so a zero outside the sensible range is part of the calculation even when it is not part of the answer.
Turn the increase into a price
Why: the increase of two dollars is added to the original ten dollars.
\[ \text{price} = 10 + 2 = 12 \]
Check by computing the revenue at the vertex and either side
Why: at an increase of two the revenue is two hundred and eighty-eight thousand dollars, and at one and at three it is lower, which confirms a maximum rather than a minimum.
\[ R(2) = -2000(-12)(12) = 288\,000 \]
\[ R(1) = -2000(-13)(11) = 286\,000, \qquad R(3) = -2000(-11)(13) = 286\,000 \]
Testing a value either side is the honest check here. It confirms both that the vertex is where you said and that it is a maximum, which the negative leading coefficient also told you.
McDougal Littell Algebra 2 (Texas Edition), Ch. 4 Quadratic Functions and Factoring — Lesson 4.4 Solve ax^2 + bx + c = 0 by Factoring §4.4, p. 262 — Example 7
Picture it
Draw the revenue function and the whole argument becomes one glance. The two zeros, the dashed axis midway between them, and the peak sitting on it.
Figure (svg): The revenue parabola for the magazine, in thousands of dollars against the price increase, crossing zero at minus ten and at fourteen with its peak of 288 thousand at an increase of two dollars, and the axis of symmetry drawn dashed.
The left zero at an increase of negative ten is the price of zero dollars: give the magazine away and the revenue is nothing, however many subscribers you have. The right zero at fourteen is the price of twenty-four dollars, at which the last subscriber leaves.
Between those two extremes the revenue has to rise and fall, and because the model is quadratic it does so symmetrically. That symmetry is the only reason averaging works, and it is why the method belongs to quadratics and not to every model.
Notice how flat the top is. Revenue at an increase of one or of three is two hundred and eighty-six thousand, only two thousand below the best, so a business getting the price slightly wrong loses very little. Near a vertex a parabola barely changes.
Read the far ends as a sanity check on the model as well. Beyond an increase of fourteen the formula predicts negative revenue, which is nonsense, and simply means the model has run out of the range it was built for.
Tweak it
The interesting quantity in the magazine problem is not the price, it is how fast subscribers leave. Change that number and the best price moves, in a direction worth predicting before you look.
Parameter explorer
Slide the number of subscribers lost per dollar of increase. Read the increase that gives the greatest revenue, and say what happens to it as the loss rate grows.
\[ R(x) = (28\,000 - {k}x)(10 + x) \]
The algebra behind the movement is worth seeing. The right-hand zero is twenty-eight thousand divided by the loss rate, so a larger loss rate pulls that zero towards zero, and the average of the zeros comes down with it.
The moment the average goes negative the advice reverses. A magazine whose readers are very price-sensitive should charge less than it currently does, and the same formula says so without any extra reasoning.
That is the real value of intercept form here. One expression covers every loss rate, so a question about a different market is a substitution rather than a fresh problem.
Worked example
The same method with the price moving the other way, which is exactly where signs get lost. A shop sells forty-five surfboards a month at five hundred dollars each, and for every twenty dollar decrease in price it sells five more.
Define the variable carefully
Why: let x be the number of twenty dollar decreases, so the price falls by twenty x and the quantity rises by five x.
\[ x = \text{number of 20 dollar decreases} \]
Write the revenue as a product
Why: quantity times price, with each bracket built from the story.
\[ R(x) = (45 + 5x)(500 - 20x) \]
Check the brackets at x equal to zero, as always. Forty-five boards at five hundred dollars is the situation described, so the model is anchored correctly before you touch it.
Take a common factor out of each bracket
Why: five comes out of the first and twenty out of the second, which makes the numbers small enough to read.
\[ R(x) = 100(9 + x)(25 - x) \]
Read the zeros and average them
Why: the brackets vanish at negative nine and at twenty-five, and the vertex is midway between.
\[ x = \dfrac{-9 + 25}{2} = 8 \]
Turn eight decreases into a price
Why: each decrease is twenty dollars, so the price falls by one hundred and sixty dollars from five hundred.
\[ \text{price} = 500 - 20(8) = 340 \]
Check the revenue at the vertex against the quantity sold
Why: eight decreases means forty-five plus forty, which is eighty-five boards, at three hundred and forty dollars each, and eighty-five times three hundred and forty is twenty-eight thousand nine hundred, which is what the factored form gives.
\[ R(8) = 100(17)(17) = 28\,900 = 85 \cdot 340 \]
The two calculations of the same revenue are the check. One uses the factored model and one counts boards and dollars, and agreement means the model and the arithmetic are both right.
McDougal Littell Algebra 2 (Texas Edition), Ch. 4 Quadratic Functions and Factoring — Lesson 4.4 Solve ax^2 + bx + c = 0 by Factoring §4.4, p. 264 — Exercise 64
Picture it
Draw this one too, because it has a feature the magazine problem did not: a zero that could never happen in the shop.
Figure (svg): The surfboard shop's monthly revenue in thousands of dollars against the number of twenty dollar price cuts, crossing zero at minus nine and at twenty-five and peaking at 28.9 thousand after eight cuts.
The zero at twenty-five decreases is a price of five hundred minus five hundred, which is zero dollars: give the boards away and there is no revenue. That one is at least imaginable.
The zero at negative nine decreases means raising the price by one hundred and eighty dollars, at which point the model predicts zero boards sold. Whether the shop would really sell none is beside the point; the average needs the number, and the model supplies it.
Notice how flat the peak is again. Seven or nine decreases give revenues of twenty-eight thousand eight hundred, only one hundred dollars below the best, so the shop has some freedom about exactly where to price.
Reading the graph also settles the direction of the answer. The peak is well to the right of zero, which says that cutting the price does raise revenue here, and a negative peak would have said the opposite.
Estimation
The book's what-if. Everything about the magazine stays the same except the starting price, which is eleven dollars rather than ten. Estimate the answer before computing it, and notice which part of the model changes.
Predict first
With twenty-eight thousand subscribers at eleven dollars, still losing two thousand per dollar of increase, what price maximises revenue and what is the maximum?
Correct: 12 dollars and 50 cents, for 312,500 dollars
Why: Only the second bracket changes, to eleven plus x, so the left-hand zero moves from negative ten to negative eleven while the right-hand zero stays at fourteen. The average of fourteen and negative eleven is one and a half, so the price is eleven plus one and a half, which is twelve dollars fifty. The revenue is twenty-five thousand subscribers at that price, which is three hundred and twelve thousand five hundred dollars.
\[ R(x) = (28\,000 - 2000x)(11 + x) = -2000(x - 14)(x + 11) \]
The estimate you should have made first is that the best price rises but by less than a dollar. Starting a dollar higher shifts one zero by one, and the average of two numbers moves by half of what one of them moves by.
Check the revenue directly rather than trusting the factored form alone. An increase of one and a half means twenty-eight thousand minus three thousand, which is twenty-five thousand subscribers, at twelve dollars fifty each, and that product is three hundred and twelve thousand five hundred.
Notice that the maximum revenue went up as well as the price. A better starting point is worth more than the price change it causes, which is the kind of conclusion this model exists to supply.
McDougal Littell Algebra 2 (Texas Edition), Ch. 4 Quadratic Functions and Factoring — Lesson 4.4 Solve ax^2 + bx + c = 0 by Factoring §4.4, p. 262 — Guided Practice 22
Pattern
Five stages, and the order is the point of the lesson. Nothing here is new; what is new is doing the cheap stages before the expensive one, which is the difference between a written search and endless guessing.
Figure (svg): A flow diagram of five stages, from taking out a common monomial factor through to expanding the answer as a check, with the reason for each stage written beside it.
Stage one and stage two cost you a glance each and can remove the search entirely. Stage three turns the remaining problem into a finite list, and stage four cuts that list roughly in half before any multiplying.
The last line is the only one that can catch a mistake, and it is the one people skip when a question is nearly finished. Expanding a pair of brackets takes ten seconds and rescues a whole answer.
One more habit worth naming: when nothing works, count what you have tried. If the list is exhausted then the trinomial does not factor over the integers, and saying so is a complete answer.
Check
All four options here multiply out to something with the correct first and last terms, or nearly so. Expand the middle of each, or use the size estimate: five times three is fifteen, and you need fourteen.
Check your understanding
Which factorisation of 5x² + 14x − 3 is correct?
Answer: D
Why: With five and one as the leading pair, the middle coefficient is five times the second constant plus the first. Pairing the five with the three and giving the one the minus sign gives fifteen minus one, which is fourteen. Expanding confirms it: five x squared plus fifteen x minus x minus three.
The option that keeps the minus sign on the one is the one to check first, because that is the arrangement the size estimate points at: you want the five multiplying the larger constant.
Two of the wrong options are the same four integers arranged differently, which is exactly the trap from Part 1. Writing the arrangement table puts both of them on the page and lets the middle column decide.
The option with a five outside the brackets is worth a second's thought as well. A common factor can only come out if it divides every coefficient, and here the constant is three.
Finish by expanding your choice. Five x squared plus fifteen x minus x minus three is five x squared plus fourteen x minus three, which is the trinomial asked about.
McDougal Littell Algebra 2 (Texas Edition), Ch. 4 Quadratic Functions and Factoring — Lesson 4.4 Solve ax^2 + bx + c = 0 by Factoring §4.4, p. 263 — Exercise 12
Check
Three of these four options expand to the right trinomial. Only one of them is fully factored, so the question is not only whether the arithmetic works but whether the job is finished.
Check your understanding
Which of these is 12x² − 28x − 24 factored completely?
Answer: A
Why: All three coefficients are divisible by four, leaving three x squared minus seven x minus six, which factors as three x plus two, times x minus three. Expanding checks it: three x squared minus nine x plus two x minus six, times four, is twelve x squared minus twenty-eight x minus twenty-four.
The two options that expand correctly and are still wrong are the interesting ones. Expanding is a necessary check and not a sufficient one, and completely means no bracket has a common factor left inside.
The quick test is to scan each bracket on its own. Twelve x plus eight has a four in every term; six x plus four has a two; three x plus two has nothing, which is what finished looks like.
The option with the signs swapped is caught by the sign rules from Part 1. The constant is negative so the two numbers disagree, and the middle coefficient is negative, so the larger product has to be the negative one.
Notice that taking the greatest common factor out at the start would have made all three wrong options impossible to write. Stage one of the procedure is a safeguard as well as a saving.
McDougal Littell Algebra 2 (Texas Edition), Ch. 4 Quadratic Functions and Factoring — Lesson 4.4 Solve ax^2 + bx + c = 0 by Factoring §4.4, p. 263 — Exercise 22
Check
A model already in intercept form, so there is no factoring left to do. Read the two zeros off the brackets, average them, and substitute.
Check your understanding
A shop's monthly revenue in dollars is R(x) = −100(x − 24)(x + 6), where x is the number of one dollar price cuts. Which cut maximises revenue, and what is the maximum?
Answer: A
Why: The brackets vanish at twenty-four and at negative six, so the axis of symmetry is at their average, which is nine. Substituting nine gives negative one hundred times negative fifteen times fifteen, which is twenty-two thousand five hundred dollars.
The wrong option that averages twenty-four and six is the one to understand rather than just avoid. The bracket x plus six is zero when x is negative six, and a sign error there moves the answer by six dollars.
One of the options names a zero of the function. Setting a bracket to zero finds where the revenue vanishes, which is the opposite of where it peaks, so a zero can never be the answer to a maximising question.
Check the winner by testing either side. At eight cuts the revenue is negative one hundred times negative sixteen times fourteen, which is twenty-two thousand four hundred, and at ten it is the same, both below the peak.
The leading coefficient tells you it is a maximum rather than a minimum. It is negative, so the parabola opens downwards and the vertex is the highest point on it.
Explain it to yourself
One idea has carried this whole lesson, and writing it in your own words is what fixes it. Answer in full sentences before reading the version below.
Discussion prompt
Explain to somebody who thinks factoring with a leading coefficient other than one is endless guessing why the search is in fact finite, and say what two things you would do first to make it as short as possible.
If your explanation used the word try more than the word list, write it again. The content of this lesson is that the candidates can be enumerated, and enumeration is not trying.
A good test of your own understanding is to predict how long a search will be. Say how many rows twenty x squared plus something x plus twelve would need, and then check yourself by writing the factor pairs of twenty and of twelve.
Try explaining the grouping route as well, in one sentence about diagonals. If you can say why the two middle pieces must multiply to the leading coefficient times the constant, you own the method rather than remembering it.
Exit ticket
One question that uses every stage of the procedure: a common factor, a search with a genuine choice in it, the zero product property, and a check.
\[ 12x^2 - 4x - 40 = 0 \]
Discussion prompt
Solve this equation. Say what you took out first and why, name the factorisation, give both solutions, and check them in the original equation.
Take the four out before anything else. Searching a leading coefficient of twelve against a constant of forty is a table of well over twenty rows, and after dividing it is a table of six.
The fractional answer is the one worth checking properly. Twenty-five ninths times twelve is one hundred over three, and the arithmetic only closes if you keep the thirds rather than rounding.
If you reached the right factorisation without writing anything down, that is fine, but write the two sign conditions in the margin anyway. They are what makes the choice a decision instead of a guess.
Last thing to notice: the answer has one whole number and one fraction, which is completely normal once the leading coefficient is not one. A denominator in a root usually comes from the leading coefficient of its bracket.
Recap
Four questions to ask of any quadratic expression, in this order, and the fourth one is the only one that ever needs a list.
| what you see | what to do | an example from this lesson |
|---|---|---|
| every term shares a factor | take the greatest one out first, sign included | 6q² − 14q + 8 = 2(3q − 4)(q − 1) |
| squares at both ends, no middle term | difference of two squares | 9x² − 64 = (3x + 8)(3x − 8) |
| squares at both ends, middle is twice the product | perfect square trinomial | 4y² + 20y + 25 = (2y + 5)² |
| none of the above | list the factor pairs, cut by signs, test the rows | 5x² − 17x + 6 = (5x − 2)(x − 3) |
| or, for a single search | two numbers with product ac and sum b, then group | 15x² − 2x − 8 = (3x + 2)(5x − 4) |
For an equation, three more moves follow: reach standard form so that one side is zero, divide out any common factor, and then set each bracket to zero and substitute both answers back into the equation you began with.
\[ R(x) = a(x - p)(x - q) \;\Longrightarrow\; \text{the extreme value is at } x = \dfrac{p + q}{2} \]
And a factored quadratic hands you its maximum or minimum for nothing, because a parabola is symmetric and its vertex sits above the midpoint of its two zeros. That is the line behind every revenue question in the exercise set.
The claim this lesson was built to defend is that the trial and error ends. Two lists, cut in half by two signs, and a handful of rows to expand; or one pair of numbers with a known product and sum. Either way you can see the bottom of the page from the top of it.
Lesson 4.5 takes the same equations and solves them by square roots instead of factoring, which finally handles the trinomials whose lists ran out. Lesson 4.8 gives the formula that solves every quadratic, and the quantity inside its square root is exactly what decides whether a list could ever have worked.
OpenStax Algebra and Trigonometry 2e, §1.5 Factoring Polynomials §1.5 — the same factoring, with the ac-method called factoring by grouping
Want this taught 1-on-1? Alexander tutors Algebra 2 — $55/session, free consultation.