2B, 2C - Quadratic roots and the discriminant

2B, 2C - Quadratic roots and the discriminant

A quadratic can have two different real solutions, a repeated real solution or no real solutions. In 2B you will use the discriminant to distinguish these cases and determine conditions on coefficients; in 2C you will find the roots exactly with the quadratic formula, ready to use expressions involving them.

The quadratic formula

Does x2−4x+5=0x^2-4x+5=0 have a real solution, or has a calculator missed one?

You already know that completing the square turns its left side into (x−2)2+1(x-2)^2+1. This is at least 11, so it cannot equal zero for real xx. We will develop a formula that both finds roots when they exist and explains this failure.

Root

A root of an equation is a value of the variable that makes the equation true. A real root of ax2+bx+c=0ax^2+bx+c=0 is also the xx-coordinate of a point where the graph y=ax2+bx+cy=ax^2+bx+c meets the xx-axis.

Completing a square produces one method for every quadratic. Starting with ax2+bx+c=0ax^2+bx+c=0, where a≠0a\ne0, multiply by 4a4a, then add the term needed for a square:

4a2x2+4abx+4ac=0multiply the equation by 4a4a2x2+4abx+b2=b2−4acadd b2 and subtract 4ac(2ax+b)2=b2−4acrecognise the square\begin{aligned} 4a^2x^2+4abx+4ac&=0 &&\text{multiply the equation by }4a\\ 4a^2x^2+4abx+b^2&=b^2-4ac &&\text{add }b^2\text{ and subtract }4ac\\ (2ax+b)^2&=b^2-4ac &&\text{recognise the square} \end{aligned}

When the right side is nonnegative, taking both signs of its square root and rearranging gives the quadratic formula.

Quadratic formula

For real coefficients with a≠0a\ne0, the real roots of ax2+bx+c=0ax^2+bx+c=0, when b2−4ac≥0b^2-4ac\ge0, are

x=−b±b2−4ac2a.x=\frac{-b\pm\sqrt{b^2-4ac}}{2a}.

The sign ±\pm means evaluate once with ++ and once with −-. If the square root is zero, these give the same value.

Not on the formulae sheet: learn it.

Put the equation in the form ax2+bx+c=0ax^2+bx+c=0 before identifying the coefficients. Keep the signs attached to bb and cc, and simplify the root before rounding anything.

Worked example 1: exact irrational roots

Question. Solve x2−4x+1=0x^2-4x+1=0, giving exact answers.

Plan. Identify a=1a=1, b=−4b=-4, c=1c=1, then substitute into the formula and simplify the surd.

x=−(−4)±(−4)2−4(1)(1)2(1)substitute signed coefficients=4±122evaluate inside the root=4±232extract the square factor 4=2±3divide both numerator terms by 2\begin{aligned} x&=\frac{-(-4)\pm\sqrt{(-4)^2-4(1)(1)}}{2(1)} &&\text{substitute signed coefficients}\\ &=\frac{4\pm\sqrt{12}}{2} &&\text{evaluate inside the root}\\ &=\frac{4\pm2\sqrt3}{2} &&\text{extract the square factor }4\\ &=2\pm\sqrt3 &&\text{divide both numerator terms by }2 \end{aligned}

Answer. x=2+3x=2+\sqrt3 or x=2−3x=2-\sqrt3.

Check. The equation is also (x−2)2−3=0(x-2)^2-3=0. Substituting either answer gives (±3)2−3=0(\pm\sqrt3)^2-3=0, so both values satisfy it exactly.

Change only the leading coefficient from 11 to 22. The denominator changes too; simplifying a surd does not remove that denominator.

Worked example 2: a nonunit leading coefficient

Question. Solve 2x2−4x+1=02x^2-4x+1=0, giving exact answers.

Plan. Use a=2a=2, b=−4b=-4, c=1c=1, retaining the denominator 2a=42a=4.

x=4±16−4(2)(1)4substitute into the formula=4±84calculate the radicand=4±224simplify the surd=1±22divide each term by 4\begin{aligned} x&=\frac{4\pm\sqrt{16-4(2)(1)}}{4} &&\text{substitute into the formula}\\ &=\frac{4\pm\sqrt8}{4} &&\text{calculate the radicand}\\ &=\frac{4\pm2\sqrt2}{4} &&\text{simplify the surd}\\ &=1\pm\frac{\sqrt2}{2} &&\text{divide each term by }4 \end{aligned}

Answer. x=1+22x=1+\frac{\sqrt2}{2} or x=1−22x=1-\frac{\sqrt2}{2}.

Check. Completing the square gives 2(x−1)2−1=02(x-1)^2-1=0. Each answer makes the left side 2(22)2−1=2×12−1=02\left(\frac{\sqrt2}{2}\right)^2-1=2\times\frac12-1=0.

Two sign mistakes can change a correct formula into the wrong calculation.

Watch out

If b=−4b=-4, then −b=4-b=4 and b2=(−4)2=16b^2=(-4)^2=16. Writing −42=−16-4^2=-16 has lost the brackets and is not the square of bb. The denominator 2a2a divides the whole numerator, including both terms.

The next roots also remain exact after simplifying a square factor.

Counting real roots

The number inside the formula's square root tells you how many different real roots are possible.

Compare three equations differing only in their constant term:

x2−4x+3=(x−2)2−1zero at x=1,3x2−4x+4=(x−2)2zero only at x=2x2−4x+5=(x−2)2+1never zero for real x\begin{aligned} x^2-4x+3&=(x-2)^2-1 &&\text{zero at }x=1,3\\ x^2-4x+4&=(x-2)^2 &&\text{zero only at }x=2\\ x^2-4x+5&=(x-2)^2+1 &&\text{never zero for real }x \end{aligned}

Their graphs show a curve crossing the axis twice, touching it once or missing it entirely.

[DIAGRAM: asset_name: Three discriminants and three root patterns; asset_slug: discriminant-root-cases; caption: Changing the constant gives two unequal real roots, equal real roots or no real roots; description: Three coordinate plots on identical scales show y equals x squared minus 4x plus 3, plus 4 and plus 5. Their discriminants are respectively 4, 0 and minus 4. The first crosses the x-axis at 1 and 3, the second touches at 2, and the third has minimum 1 at x equals 2 and never meets the x-axis.]
Changing the constant gives two unequal real roots, equal real roots or no real roots

The algebra explains all three pictures.

Discriminant

The discriminant of ax2+bx+c=0ax^2+bx+c=0, where a≠0a\ne0, is

D=b2−4ac.D=b^2-4ac.

It is the expression under the square root in the quadratic formula.

  • If D>0D>0, then D>0\sqrt D>0: adding and subtracting it gives two unequal real roots.
  • If D=0D=0, the formula gives x=−b2ax=-\frac b{2a} twice: equal real roots, also called a repeated root. There is one distinct value, and the curve touches the axis there.
  • If D<0D<0, no real number has square DD: there are no real roots, and the curve does not meet the axis.

You do not need to calculate the roots just to classify them.

Worked example 3: equal real roots

Question. Determine the nature of the roots of x2−4x+4=0x^2-4x+4=0.

Plan. Calculate the discriminant and interpret its sign.

D=(−4)2−4(1)(4)use a=1, b=−4, c=4=16−16=0compare the two terms\begin{aligned} D&=(-4)^2-4(1)(4) &&\text{use }a=1,\ b=-4,\ c=4\\ &=16-16=0 &&\text{compare the two terms} \end{aligned}

Answer. The roots are equal and real because D=0D=0.

Check. Factorising gives (x−2)2=0(x-2)^2=0. Both factors give x=2x=2, one distinct repeated root.

Increase only the constant from 44 to 55. The discriminant moves below zero, so the repeated root disappears from the real number line.

Worked example 4: no real roots

Question. Determine the nature of the roots of x2−4x+5=0x^2-4x+5=0.

Plan. Evaluate the discriminant; a negative value rules out real roots.

D=(−4)2−4(1)(5)keep the negative coefficient in brackets=16−20=−4calculate exactlyD<0interpret the sign\begin{aligned} D&=(-4)^2-4(1)(5) &&\text{keep the negative coefficient in brackets}\\ &=16-20=-4 &&\text{calculate exactly}\\ D&<0 &&\text{interpret the sign} \end{aligned}

Answer. There are no real roots.

Check. The left side is (x−2)2+1≥1(x-2)^2+1\ge1, so it cannot be zero for real xx. This independently confirms the conclusion.

A conclusion matters as much as the calculation that supports it.

Exam tip

Write what the sign means: for example, “D=−4<0D=-4<0, so there are no real roots.” A calculator error message alone does not explain why a real root is impossible.

Classify the next equation without calculating its roots.

Conditions on coefficients

A coefficient containing an unknown constant turns root classification into a condition on that constant.

A parameter is a constant whose value may vary between equations. For instance, changing kk in x2−6x+k=0x^2-6x+k=0 gives a family of quadratics. Calculate the discriminant in terms of the parameter, then use D>0D>0, D=0D=0 or D<0D<0 according to the roots requested.

The boundary D=0D=0 belongs to equal roots. Strict inequalities are needed for unequal roots and for no real roots. Also check the leading coefficient: if it becomes zero, the equation is no longer quadratic and the discriminant rule does not classify it.

Worked example 5: a parameter in the constant term

Question. For x2−6x+k=0x^2-6x+k=0, find the values of kk giving equal real roots, unequal real roots and no real roots.

Plan. Express the discriminant in terms of kk, then solve each sign condition.

D=(−6)2−4(1)(k)=36−4kidentify the coefficientsD=0  ⟺  k=9the boundary caseD>0  ⟺  36>4k  ⟺  k<9two unequal real rootsD<0  ⟺  36<4k  ⟺  k>9no real roots\begin{aligned} D&=(-6)^2-4(1)(k)=36-4k &&\text{identify the coefficients}\\ D=0&\iff k=9 &&\text{the boundary case}\\ D>0&\iff 36>4k\iff k<9 &&\text{two unequal real roots}\\ D<0&\iff 36<4k\iff k>9 &&\text{no real roots} \end{aligned}

Answer. Equal real roots: k=9k=9. Unequal real roots: k<9k<9. No real roots: k>9k>9.

Check. At the boundary, the expression is (x−3)2(x-3)^2. On either side, k=8k=8 gives (x−2)(x−4)(x-2)(x-4), with two different roots, while k=10k=10 gives (x−3)2+1(x-3)^2+1, which is never zero. The equivalent equation (x−3)2=9−k(x-3)^2=9-k confirms all three conditions.

Now move the parameter from the constant term to the leading coefficient, leaving the other two coefficients as −6-6 and 11. The discriminant happens to be the same, but k=0k=0 now needs separate treatment.

Worked example 6: checking when the equation stops being quadratic

Question. Classify the roots of kx2−6x+1=0kx^2-6x+1=0 in terms of kk. Deal separately with k=0k=0.

Plan. Apply the discriminant only when k≠0k\ne0, then substitute k=0k=0 into the original equation.

For k≠0k\ne0,

D=(−6)2−4(k)(1)=36−4know a=kD=0  ⟺  k=9equal real rootsD>0  ⟺  k<9retain the condition k≠0D<0  ⟺  k>9no real roots\begin{aligned} D&=(-6)^2-4(k)(1)=36-4k &&\text{now }a=k\\ D=0&\iff k=9 &&\text{equal real roots}\\ D>0&\iff k<9 &&\text{retain the condition }k\ne0\\ D<0&\iff k>9 &&\text{no real roots} \end{aligned}

At k=0k=0, the original equation becomes −6x+1=0-6x+1=0, so x=16x=\frac16. It is linear, not a quadratic with repeated roots.

Answer. Equal real roots for k=9k=9; unequal real roots for k<9k<9 with k≠0k\ne0; no real roots for k>9k>9. At k=0k=0, there is the single linear solution x=16x=\frac16.

Check. At k=9k=9, the left side is (3x−1)2(3x-1)^2. At k=8k=8, it is (4x−1)(2x−1)(4x-1)(2x-1), giving different roots 14\frac14 and 12\frac12. At k=10k=10, it is 10(x−310)2+110>010(x-\frac3{10})^2+\frac1{10}>0. Substitution of k=0k=0, x=16x=\frac16 gives −1+1=0-1+1=0.

The restriction comes from the equation's degree, not from solving the inequality.

Watch out

For kx2−6x+1=0kx^2-6x+1=0, the inequality D>0D>0 alone would include k=0k=0. Exclude it before claiming two quadratic roots. The formula assumes a nonzero leading coefficient.

The same check applies whenever a parameter can remove the squared term.

Quick Recap

  • Put a quadratic in the form ax2+bx+c=0ax^2+bx+c=0, with a≠0a\ne0, before using the formula.
  • Keep both signs and simplify exact surds; divide the whole numerator by 2a2a.
  • The sign of D=b2−4acD=b^2-4ac distinguishes unequal real roots, equal real roots and no real roots.
  • For parameter conditions, justify strict endpoints and check whether the leading coefficient can vanish.

The opening equation has D=−4D=-4, so it has no real solution. Its completed square (x−2)2+1(x-2)^2+1 explains the same result geometrically.