Relvora Prep

GRE Quant lesson · Inequalities

GRE inequalities: flipping signs, ranges and quadratic inequalities

Inequalities for GRE Quant: solving linear and compound inequalities, flipping when multiplying by a negative, combining ranges, and quadratic sign charts.

The core idea

Inequalities solve like equations with one extra rule: multiplying or dividing by a negative number flips the sign. To find the extreme value of an expression, push each variable to the end of its range that helps; for quadratics, factor and check the sign between and outside the roots.

Core ideas

Words used here

  • Inequality: a statement that one side is smaller or larger: < (less than), ≤ (less than or equal to), > (greater than), ≥ (greater than or equal to).
  • Strict inequality: < or >. The two sides can't be equal, so the boundary value is not included.
  • Boundary (or endpoint): the value where the two sides are equal. It is the end of a range.
  • Compound inequality: an inequality with three parts, such as −4 < 2x + 6 ≤ 10.
  • Flip: reverse the direction of the sign, so < becomes > and ≤ becomes ≥.
  • Quadratic: an expression with x² in it, such as x² − x − 12.
  • Root: a value of x that makes an expression equal to 0.
  • Integer: a whole number, such as −3, 0 or 8.
  • Counterexample: one value that makes a statement false.
  • Adding or subtracting the same number on both sides keeps the direction of the sign.
  • Multiplying or dividing both sides by a positive number keeps the direction. By a negative number, flip the sign. Why: 2 < 5 is true. Multiply both sides by −1 and you get −2 and −5. Now −2 is the larger number, so −2 > −5.
  • Compound inequalities: do each step to all three parts.
  • Largest or smallest value of an expression: try the ends of each range. For a product, try every pair of ends.
  • Quadratics: factor, find the roots, and test the sign in each region around them.

On the GRE

  • The on-screen calculator doesn't decide which way a sign points; reasoning does.
  • Quantitative Comparison: subtract the quantities and ask whether the difference keeps one sign for every allowed value. Try a negative value, a fraction and an endpoint before you trust one test.
  • Select-all: "Which of the following could be the value of x? Indicate all such values." Solve the inequality once, then check every choice, endpoints included or excluded as the signs say.

Try it. Solve 4x − 7 ≥ 9.

Show the answer
  1. Add 7 to both sides: 4x ≥ 16.
  2. Divide both sides by 4 (a positive number, so no flip): x ≥ 4.
  3. Check: x = 4 gives 16 − 7 = 9, and 9 ≥ 9 ✓.

Solving a linear inequality

Get the variable alone, as you would in an equation. In a compound inequality (three parts), do every step to all three parts.

Why it works. Adding or subtracting the same number slides both sides the same distance along the number line, so their order can't change. Dividing by a positive number shrinks or stretches the line without turning it around, so the order is kept as well.

How to spot it: a linear expression (no x²) with <, ≤, > or ≥, often with "how many integers".

Example. How many integers satisfy −4 < 2x + 6 ≤ 10?

  1. Subtract 6 from all three parts: −10 < 2x ≤ 4.
  2. Divide all three parts by 2: −5 < x ≤ 2.
  3. The < means −5 is not included. The ≤ means 2 is included.
  4. Integers: −4, −3, −2, −1, 0, 1, 2. Count: last − first + 1 = 2 − (−4) + 1 = 7.

Number line from −6 to 3. The range from −5 to 2 is shaded. An open circle at −5 shows that −5 is not included; a filled dot at 2 shows that 2 is included.

On a number line, an open circle marks an endpoint that is not included (< or >). A filled dot marks an endpoint that is included (≤ or ≥).

Trap: including −5, which the strict < leaves out.

Try it on the number line. Move the two ends and choose whether each one is included. The readout counts the integers in the range.

Number line

−5 < x ≤ 2.

So −5 < x ≤ 2. Open dot at −5, filled dot at 2.

Integers in it: −4, −3, −2, −1, 0, 1, 2, which is 2 − (−4) + 1 = 7 integers.

Edge cases.

  • Ends that aren't integers. −2.5 < x ≤ 3.5 contains the integers −2 to 3. Find the first and last integers inside the range before you use last − first + 1: 3 − (−2) + 1 = 6.
  • An empty range. 5 < x < 5 contains no numbers at all.
  • "And" vs. "or". "x > 2 and x < 7" is one range, 2 < x < 7. "x < 2 or x > 7" is two separate pieces.

GRE-level example. How many integers n satisfy both 3n − 5 < 2n + 4 and 7 − 2n ≤ 3n + 22?

  1. First inequality: subtract 2n from both sides and add 5: n < 9.
  2. Second inequality: add 2n to both sides and subtract 22: 7 − 22 ≤ 5n, so −15 ≤ 5n and n ≥ −3.
  3. Both together: −3 ≤ n < 9. So −3 is included and 9 is not.
  4. The integers are −3 to 8: 8 − (−3) + 1 = 12.
  5. Check the ends: n = −3 gives 7 + 6 = 13 ≤ −9 + 22 = 13 ✓. n = 9 gives 22 < 22, which is false ✓.
both: −3 ≤ n < 9n < 9n ≥ −3−5−4−3−2−1012345678910
Only the overlap counts: the integers from −3 to 8, which is 12 of them

Check yourself. How many integers x satisfy −7 ≤ 3 − 2x < 5?

Check your answer

Subtract 3 from all three parts: −10 ≤ −2x < 2. Divide all three parts by −2 and flip both signs: 5 ≥ x > −1, which is −1 < x ≤ 5. The integers are 0 to 5: 6.

Flipping when multiplying by a negative

When you multiply or divide both sides by a negative number, flip the sign: < becomes >, and > becomes <. Check with a test value.

Why it works. Multiplying by −1 mirrors every number across 0. A number that was to the right of another ends up to its left, so the order reverses. Multiplying by any negative number is that mirror plus a stretch, so it reverses the order too.

How to spot it: a negative coefficient (the number in front of the variable, like −3 in −3x), or an expression divided by a negative number.

Example. Solve 5 − 3x > 17.

  1. Subtract 5 from both sides: −3x > 12.
  2. Divide both sides by −3. The number becomes 12 ÷ (−3) = −4, and the sign flips from > to <: x < −4.
  3. Check x = −5, which should work: 5 − 3(−5) = 5 + 15 = 20, and 20 > 17 ✓.
x < −4trap: x > −4−8−7−6−5−4−3−2−1012test x = −5 ✓
Dividing by −3 flips > to <, so the solutions lie to the left of −4

Trap: keeping the sign and writing x > −4.

Try it. Solve 3 − 2x ≤ 11.

Show the answer
  1. Subtract 3 from both sides: −2x ≤ 8.
  2. Divide both sides by −2 and flip ≤ to ≥: x ≥ 8 ÷ (−2), so x ≥ −4.
  3. Check x = 0: 3 ≤ 11 ✓. Check x = −5, which should fail: 3 + 10 = 13, and 13 ≤ 11 is false ✓.

Edge cases.

  • Never multiply by a variable whose sign you don't know. To solve 6/x > 2, you can't simply write 6 > 2x. If x > 0, you get 6 > 2x, so x < 3. If x < 0, you must flip: 6 < 2x, so x > 3, which contradicts x < 0. So the answer is 0 < x < 3.
  • Compound inequalities flip both signs. −6 < −2x < 4 becomes 3 > x > −2.
  • Reciprocals flip too. If a and b are both positive and a < b, then 1/a > 1/b (2 < 4, but 1/2 > 1/4).

GRE: Quantitative Comparison. Information: x > y. Quantity A: 5 − 2x. Quantity B: 5 − 2y.

  1. Multiply x > y by −2 and flip: −2x < −2y. Add 5 to both sides: 5 − 2x < 5 − 2y.
  2. Quantity B is greater: (B).
  3. Compare Quantity A: x with Quantity B: 2y instead. x = 3, y = 2 makes B greater, but x = −1, y = −2 makes A greater: (D). The difference x − 2y still depends on y.

GRE: select-all. If −7 < 5 − 3x ≤ 8, which of the following could be x? Subtract 5: −12 < −3x ≤ 3. Divide by −3 and flip both signs: 4 > x ≥ −1. So select every choice from −1 (included) up to, but not including, 4.

GRE-level example. If (4 − x)/(−3) ≥ 2, what is the least possible value of x?

  1. Multiply both sides by −3. It's negative, so ≥ becomes ≤: 4 − x ≤ −6.
  2. Subtract 4 from both sides: −x ≤ −10.
  3. Multiply by −1 and flip again: x ≥ 10.
  4. The least possible value is 10.
  5. Check: x = 10 gives (−6)/(−3) = 2, and 2 ≥ 2 ✓. x = 9 gives (−5)/(−3) = 5/3, which is less than 2 ✓ (9 fails, as it should).

Check yourself. Solve −x/2 + 1 > 4.

Check your answer

Subtract 1: −x/2 > 3. Multiply by −2 and flip: x < −6. Check x = −8: 4 + 1 = 5 > 4 ✓.

Combining ranges for a maximum or minimum

Push each variable to the end of its range that helps. For a sum or a difference, that's one choice per variable. For a product, test all four pairs of endpoints, because two negative numbers multiply to a positive one.

Why it works. a − b grows when a grows or when b shrinks, so the greatest difference uses the largest a and the smallest b. For a product, which end helps depends on the signs: a large negative times a large negative is a large positive. For a fixed b, a × b changes steadily as a moves, so its largest and smallest values sit at the ends of a's range. The same is true for b. That's why the four endpoint pairs can't miss the extremes.

How to spot it: "if a ≤ x ≤ b and c ≤ y ≤ d, what is the greatest (or least) value of…".

Example. If −2 ≤ a ≤ 3 and −5 ≤ b ≤ 1, what are the least and greatest values of ab?

  1. Multiply each pair of endpoints: (−2)(−5) = 10, (−2)(1) = −2, (3)(−5) = −15, (3)(1) = 3.
  2. Least: −15. Greatest: 10.
ab runs from −15 to 10−16−14−12−10−8−6−4−2024681012(3)(−5) = −15(−2)(1) = −2(3)(1) = 3(−2)(−5) = 10
The four endpoint products: the extremes are −15 and 10, not the product of the two maximums, 3

For a difference with the same ranges, the greatest a − b uses the largest a and the smallest b: 3 − (−5) = 3 + 5 = 8.

Trap: multiplying the two maximums (3 × 1 = 3) and calling it the greatest.

Edge cases.

  • A square isn't steady. If −3 ≤ a ≤ 2, then a² runs from 0 (at a = 0, inside the range) up to 9 (at a = −3). The least value is not at an end.
  • Strict ends. With −2 < a < 3, the value 3 is never reached, so "the greatest possible value of a" doesn't exist. GRE questions then use ≤, or ask about integers.
  • Integers only. If a and b must be integers, use the largest and smallest integers inside each range.

GRE: Quantitative Comparison. Information: −2 < x < 5 and 1 < y < 3. Quantity A: x − y. Quantity B: 4.

  1. x − y is largest with the largest x and the smallest y: just under 5 − 1 = 4. It never reaches 4, because both inequalities are strict.
  2. So Quantity B is always greater: (B).
  3. If Quantity B were 2, values such as x = 4, y = 1.5 (x − y = 2.5) and x = 0, y = 2 (x − y = −2) fall on both sides: (D).

GRE-level example. If −3 ≤ x ≤ 2 and 1 ≤ y ≤ 4, what is the greatest possible value of x² − y?

  1. To make x² − y large, make x² as large as possible and y as small as possible.
  2. x² is largest at the end farthest from 0: x = −3 gives 9. (x = 2 gives only 4.)
  3. The smallest y is 1.
  4. Greatest value: 9 − 1 = 8.

The trap answer 3 uses the largest x, 2: 2² − 1 = 3.

Check yourself. If 2 ≤ a ≤ 5 and −4 ≤ b ≤ −1, what is the least possible value of a/b?

Check your answer

Try the four pairs: 2/(−4) = −0.5, 2/(−1) = −2, 5/(−4) = −1.25, 5/(−1) = −5. The least is −5: the largest a over the b closest to 0.

Quadratic sign chart

Move everything to one side so the other side is 0. Factor the quadratic (the expression with x²) and find its roots (the x-values that make it 0). The roots split the number line into regions; test one value in each. When the number in front of x² is positive, the expression is positive outside the roots and negative between them.

Why it works. A product can change sign only where one of its factors passes through 0. So between two roots the sign of (x − 4)(x + 3) can't change, and one test value gives the sign of the whole region. For large x both factors are positive; each root you pass going left flips one factor, so the signs alternate: +, −, +.

How to spot it: x² terms in an inequality.

Example. Solve x² − x − 12 > 0.

  1. Factor: find two numbers that multiply to −12 and add to −1. They are −4 and 3. So (x − 4)(x + 3) > 0.
  2. Roots: x − 4 = 0 gives x = 4, and x + 3 = 0 gives x = −3.
  3. Test one value in each region:
Region x < −3 −3 < x < 4 x > 4
Test value −4 0 5
(x − 4)(x + 3) (−8)(−1) = 8 (−4)(3) = −12 (1)(8) = 8
Positive? yes no yes
  1. Answer: x < −3 or x > 4.
+ (test −4: 8)− (test 0: −12)+ (test 5: 8)−6−5−4−3−2−101234567root −3root 4
(x − 4)(x + 3) is positive outside the roots and negative between them, so x < −3 or x > 4

For a single square, x² ≤ 30: 5² = 25 works, but 6² = 36 is too big. Negatives work too, because (−5)² = 25. So every integer from −5 to 5 works: eleven integers, including 0.

Trap: solving only x − 4 > 0 and missing x < −3.

See it. The graph of y = x² − x − 12 is above the x-axis (positive) outside the roots −3 and 4, and below it (negative) between them. Change a, b and c to see the regions move; with a negative, the pattern turns upside down.

Parabola: roots and vertex

y = x² − x − 12. It opens up (a > 0). Vertex (1/2, −49/4), on the line x = 1/2.

Discriminant b² − 4ac = 49.

Roots x = −3 and x = 4. Sum = −b/a = 1, product = c/a = −12.

y < 0 between the roots (−3 < x < 4), and y > 0 outside them.

Edge cases.

  • A double root. (x − 2)² > 0 is true for every x except 2. (x − 2)² < 0 is never true. (x − 2)² ≤ 0 is true only at x = 2.
  • No real roots. x² + 4 > 0 is true for every x, because x² ≥ 0.
  • A negative x² coefficient. For −x² + x + 12 > 0, multiply by −1 and flip: x² − x − 12 < 0, so −3 < x < 4.
  • Don't divide by a factor. Dividing (x − 4)(x + 3) > 0 by (x − 4) assumes you know its sign. You don't.

GRE: Quantitative Comparison. Information: x² − x − 12 < 0. Quantity A: x. Quantity B: −3.

  1. Factor: (x − 4)(x + 3) < 0, which holds between the roots: −3 < x < 4.
  2. Every allowed x is greater than −3, so the answer is (A). (With ≤ instead of <, x could equal −3, and the answer would be (D).)
  3. For x² − x − 12 > 0, the solutions are x < −3 or x > 4: x can be very small or very large, so comparing x with any number is (D).

GRE-level example. How many integers n satisfy both (n + 5)(n − 3) ≤ 0 and n² > 4?

  1. (n + 5)(n − 3) ≤ 0: the roots are −5 and 3, and the product is negative between them. With ≤, the roots count too: −5 ≤ n ≤ 3.
  2. n² > 4: the numbers farther than 2 from 0, so n < −2 or n > 2.
  3. Both at once: −5 ≤ n < −2, or 2 < n ≤ 3.
  4. The integers are −5, −4, −3 and 3: 4.
  5. Check n = −2: (3)(−5) = −15 ≤ 0 ✓, but (−2)² = 4 is not greater than 4 ✓ (−2 is out).
(n + 5)(n − 3) ≤ 0n² > 4n² > 4−7−6−5−4−3−2−1012345
Both conditions hold only at −5, −4, −3 and 3: 4 integers

Check yourself. Solve x² < 4x.

Check your answer

Move 4x: x² − 4x < 0, so x(x − 4) < 0. The product is negative between the roots: 0 < x < 4. Dividing both sides by x would give x < 4, which wrongly includes negative numbers (x = −1 gives 1 < −4, false).

Must be true for a range

Test each statement with numbers from the range, including values near the ends. A counterexample is one value that makes a statement false. One counterexample shows that a statement is not always true, so you can cross it out. A statement that passes your tests still needs a reason that works for every value in the range.

Why it works. One counterexample is enough to prove "not always true". But passing a few tests proves nothing: a "must be true" statement needs a reason that covers every value in the range. For powers and reciprocals, the reason is almost always the sign of x and whether x is bigger or smaller than 1 in size. This table holds for every x in each region:

Region Example Order, smallest to largest
x > 1 x = 2 1/x < x < x² < x³ (1/2, 2, 4, 8)
0 < x < 1 x = 1/2 x³ < x² < x < 1/x (1/8, 1/4, 1/2, 2)
−1 < x < 0 x = −1/2 1/x < x < x³ < x² (−2, −1/2, −1/8, 1/4)
x < −1 x = −2 x³ < x < 1/x < x² (−8, −2, −1/2, 4)

How to spot it: "if −1 < x < 0, which of the following must be true?" On the GRE this is usually a select-all question ending "Indicate all such statements": decide each statement on its own and select every one that is always true. The examples below label the statements I, II and III to keep track of them.

Example. If −1 < x < 0, which must be true? I. x² < x. II. x³ > x. III. 1/x < −1.

  1. Test x = −1/2: x² = 1/4, x³ = −1/8, and 1/x = −2.
  2. I: is 1/4 < −1/2? No, a positive number is never less than a negative one. I is false.
  3. II: is −1/8 > −1/2? Yes. Reason: x³ = x × x², and x² is between 0 and 1, so x³ is closer to 0 than x. For negative numbers, closer to 0 means larger.
  4. III: is −2 < −1? Yes. Reason: x is between −1 and 0, so 1/x is negative with a size above 1 (for example, 1/(−0.1) = −10).
  5. Answer: II and III.

Trap: testing only positive values, where the rules for powers work the other way.

Try it. If 0 < x < 1, which must be true? I. x² < x. II. √x < x. III. 1/x > 1.

Show the answer
  1. Test x = 1/4: x² = 1/16, √x = 1/2, 1/x = 4.
  2. I: 1/16 < 1/4 ✓. Reason: x² = x × x, and multiplying by a number less than 1 makes x smaller.
  3. II: 1/2 < 1/4? No. This counterexample shows II is not always true.
  4. III: 4 > 1 ✓. Reason: dividing 1 by a number less than 1 gives more than 1.
  5. Answer: I and III.

Edge cases.

  • Test the ends when they are included. If −1 ≤ x ≤ 0, try x = 0 (then 1/x doesn't exist) and x = −1 (then x³ = x, so x³ < x fails).
  • "Could be true" is the opposite task. One example where the statement works is enough.

GRE-level example. If x < −1, which must be true? I. x³ < x II. 1/x > x III. x² > −x

  1. Test x = −2: x³ = −8, 1/x = −1/2, x² = 4, −x = 2. All three statements hold. Now find a reason for each.
  2. I: x³ = x × x², and x² > 1. Multiplying the negative number x by more than 1 moves it further left. So x³ < x always ✓.
  3. II: 1/x is between −1 and 0, while x is below −1. So 1/x > x always ✓.
  4. III: −x = |x|, which is bigger than 1. A number bigger than 1 is smaller than its square, so x² = |x|² > |x| = −x always ✓.
  5. Answer: I, II and III.

Check yourself. If 0 < x < 1, put x, x², √x and 1/x in order from smallest to largest.

Check your answer

x² < x < √x < 1/x. With x = 1/4: 1/16 < 1/4 < 1/2 < 4. Squaring a number between 0 and 1 makes it smaller; taking its square root makes it bigger, but still below 1.

How hard questions disguise it

For harder questions (L4–L5)

L4 versions ask how many integers satisfy two inequalities at once, or hide the inequality in a story: "at most 12" means ≤ 12, and "no fewer than 5" means ≥ 5. At L5, inequalities combine with absolute value (|x − 2| < 5 is a range), with optimization (the largest possible value under some conditions) and with exponents (for 0 < x < 1, higher powers are smaller). Test boundary values and fractions before you trust your first idea.

Practice

Practice inequality patterns in timed sets on the inequalities practice page.

Common traps

  • Pairing a maximum with a maximum

    To make x − y large, x must be large and y small; for a product, two large negatives can beat two positives. Test every endpoint pair when signs can differ.

  • Answering the reverse difference or slipping a sign

    x − y and y − x have opposite signs, subtracting a negative endpoint adds, and a negative times a negative is positive. Write the target expression with the numbers in parentheses.

  • Including a boundary that a strict inequality excludes

    At a boundary the two sides are equal, which fails < and >. Strict inequalities exclude the boundary values, and 'positive' excludes 0.

  • Dropping a boundary that ≤ or ≥ includes

    ≤ and ≥ include the values that make the two sides equal. Test each endpoint before you count.

  • Counting the wrong region

    An upward-opening quadratic is negative only between its roots, and when two conditions apply, only their overlap counts. Test one value in each region instead of guessing.

  • Reading x² > 4 as x > 2 only

    Squares hide the sign: x² > 4 also holds for x < −2, and x² ≤ 9 includes −3 through 3. Count both sides of zero.

  • Forgetting to flip, or flipping the numbers too

    Dividing by a negative flips the inequality sign, not the sign of the number you get. Do the arithmetic carefully, then reverse only the symbol.

  • Stopping before the variable is alone, or answering a different question

    Get the variable completely alone, and keep everything on the right side. Then reread the question: it may want the sum of the solutions, not how many there are.

  • Multiplying by a variable of unknown sign

    If you don't know whether x is positive, you can't multiply both sides by x without risking a flip. Move everything to one side and analyze signs instead.

  • Removing only half of a band around zero

    x² > c fails for every x from −√c to √c, on both sides of zero. List the window from the first condition, ends included, then remove that whole band.

  • Forgetting to flip, or misjudging an endpoint

    Dividing by −3 reverses every sign in the inequality. A strict sign leaves its endpoint out; ≤ or ≥ keeps it in.

  • Pairing the wrong extremes, or reading a quadratic's sign wrong

    The smallest x − y uses the smallest x and the largest y. (x − r)(x − s) < 0 between the roots and > 0 on both sides outside them. A value must satisfy every condition given.