The graphs of and cross at exactly one point. Solving the equation by squaring, though, produces two numbers, and . Only one of them is a crossing point. The other was made by the algebra, and the steps that make such false answers, clearing fractions and squaring, are exactly the steps these equations need.
Key terms
Original domain
Definition The original domain of an equation in a real variable is the set of real numbers at which every expression in the equation, on both sides, is defined as a real number. For rational and radical equations this means that no denominator is and that the radicand of every even root (square root, fourth root and so on) is nonnegative. Odd roots such as add no restriction. A solution of the equation is a number in the original domain that makes the two sides equal.
In plain words Before any algebra, write down which numbers are allowed, the same way you find a natural domain. A number outside that set can never be a solution, whatever the later steps say.
Example Consider the equation
The square root needs , so , and the fraction needs . The original domain is . If the algebra later produces or , that number is rejected at once.
Equivalent transformations
Definition Two equations are equivalent on a set if they have exactly the same solutions in . A step that turns an equation into one that is equivalent to it on the original domain is an equivalent transformation. The following steps are equivalent transformations.
- Adding the same expression to both sides, or subtracting it, when the expression is defined at every point of .
- Multiplying or dividing both sides by the same nonzero number.
- Multiplying both sides by an expression that is defined and nonzero at every point of .
Equivalence only speaks about numbers in . The new equation may also be defined and true at numbers outside , and those numbers are never solutions. Other common steps are not always equivalent. Multiplying both sides by an expression that is at some point of , or squaring both sides, can add solutions. Dividing both sides by an expression that is at some point of can lose solutions.
In plain words Safe steps keep the solution set exactly as it was. After a risky step you hold candidates, not answers, and you have to sort them.
Example The equation has two solutions. Subtracting is safe and gives , so the solutions are and . Dividing both sides by instead gives , and the solution disappears, because dividing by is not allowed when .
Least common denominator
Definition Factor every denominator completely. The least common denominator (LCD) of several rational expressions is the product of the least common multiple of the numerical factors and every distinct polynomial factor that appears, each raised to the highest power with which it occurs in any one denominator. Multiplying both sides of a rational equation by the LCD clears every fraction. The LCD is nonzero at every point of the original domain, so on that domain this is an equivalent transformation. The cleared equation, however, is defined for all real , and it may have solutions where the LCD is , which lie outside the original domain.
In plain words Use the smallest expression that every denominator divides into. A larger common denominator also clears the fractions, but it raises the degree and brings in extra candidates to reject.
Example Solve the equation
Since , the original domain is and , and the LCD is . Multiplying both sides by it gives
The candidate is in the domain and checks, since
Multiplying by the larger common denominator instead leads to , with the extra candidates and , both outside the domain.
Squaring
Definition For real numbers and , if then , but only says that or . So squaring both sides of keeps every solution and may add the solutions of . When and for every under consideration, squaring is an equivalent transformation. For an isolated square root this gives an exact rule,
The condition need not be added, because already forces it.
In plain words Isolate the root first, then square. Squaring cannot tell from , so it also solves the equation with the opposite sign, and those extra answers must be weeded out. If a second root is left, isolate it and square again.
Example To solve , square both sides.
The candidates are and . The condition keeps and rejects , and indeed . The number solves instead. On a graph, the squared equation finds where the line meets the whole sideways parabola . The upper half gives the point and the lower half gives .
Extraneous solutions
Definition An extraneous solution is a number that solves an equation reached later in the work but does not solve the original equation. Extraneous solutions have two sources. A step that is not an equivalent transformation, mainly squaring both sides, can add numbers inside the original domain. And a later equation may be defined at numbers outside the original domain and hold there. Clearing fractions does this, because the cleared equation is defined for every real and may hold where the LCD is , and those are exactly the numbers a denominator excludes. Extraneous solutions are not arithmetic mistakes, so redoing the algebra carefully will not make them go away.
In plain words An extraneous solution is a false answer that the method itself creates. It satisfies the equation you ended with but not the one you started with.
Example Solve the equation
The original domain is and . Multiplying by the LCD gives
The candidate is extraneous, since it lies outside the original domain. The candidate checks, because
The only solution is .
Checking candidates in the original equation
Definition Once the work includes squaring, clearing fractions or any other step that can bring in extraneous solutions, the numbers you find at the end are only candidates. A candidate is a solution exactly when lies in the original domain and substituting into the original equation, not into a later one, makes both sides equal. Substituting into the equation where a candidate appeared cannot expose it, because every candidate satisfies that equation. When the equation is and a candidate satisfies , testing is an exact check. This helps when is irrational, where a decimal substitution only gives approximate evidence.
In plain words Go back to the very first line and test each candidate there. Checking in the squared or cleared equation proves nothing, since the false answers pass that test too.
Example Solve . Isolating the root gives , and squaring gives
| Candidate | Squared equation | Original equation |
|---|---|---|
| , true | , true | |
| , true | , false |
Both candidates pass the squared equation, and only passes the original one, so the only solution is . The sign test agrees, since while .
The sign test is handy for irrational candidates. Squaring gives , with candidates
Only the positive one, , satisfies , so it is the only solution. The other candidate, about , lies in the original domain , so the domain alone does not reject it. The sign test does.
Check yourself
- Solve the equation
Show the answer
The original domain is and . Multiplying by the LCD gives , so and . That candidate is outside the original domain, so the equation has no solution.
- Solve .
Show the answer
Squaring both sides gives
At the left side is and the right side is , so is extraneous. At both sides equal . The only solution is .
- Solve .
Show the answer
The original domain is . Isolate one root as and square.
Squaring again gives , so and the candidates are and . Both check, since and . On the domain both sides were nonnegative before each squaring, so each squaring was an equivalent transformation there and every candidate in the domain had to be a solution. Dividing by would have lost the solution .