To solve an inequality, isolate the variable with the same inverse operations you use for equations, then reverse the inequality sign whenever you multiply or divide both sides by a negative number. To graph it, plot the boundary point on a number line (open circle for < or >, closed circle for ≤ or ≥) and shade the region that makes the statement true. Systems require shading overlap; two-variable inequalities require testing a point on a coordinate plane.
The core mechanism: order versus equality
An equation asserts that two expressions are identical. An inequality asserts an order relationship, and order has a property equations lack: it flips under negation. If 3 < 7, multiplying both sides by −1 gives −3 > −7. The equality survives sign changes; the inequality does not.
This single fact explains nearly every procedural difference between solving equations and solving inequalities. Addition and subtraction preserve order unconditionally. Multiplication and division preserve order only when the multiplier is positive. Multiplying or dividing by a negative quantity reverses the direction of the comparison, because the number line is being reflected through zero.
Function behavior explains why the rule is narrower than many students expect. Squaring both sides is not a valid operation on inequalities in general: −5 < 3, but squaring gives 25 > 9, a reversal triggered not by a negative multiplier but by the non-monotonic shape of the squaring function. Taking reciprocals also reverses order when both sides share a sign. The reliable operations are the monotonic ones.
Plain language: addition slides points along the number line without reordering them; negation mirrors the line, and mirroring reverses left and right.
Boundary notation: open, closed, and the infinity exception
The circle convention on a number line encodes whether the boundary satisfies the inequality. For x > 2, the point 2 is excluded, so the graph uses an open circle and interval notation (2, ∞). For x ≥ 2, the point is included: closed circle, interval [2, ∞).
Infinity is never bracketed. The symbols ∞ and −∞ describe unboundedness rather than numbers, so they always take parentheses: (−∞, 3] or (2, ∞). A reported solution like [2, ∞] is never correct.
A professional caveat worth internalizing: interval notation, set-builder notation, and inequality notation all describe the same solution set, and grading standards vary by course and country. The mathematics is identical; only the encoding differs. When a solution is ambiguous, restate it in inequality form, which is universally readable.
Quantified context: where inequality solving actually gets used
Inequality methods underpin decision thresholds across applied fields. A quality-control rule that a component must weigh within a tolerance band is a compound inequality. A budget constraint is linear. A dosage window, a minimum tensile strength, a maximum latency requirement — each is a boundary condition expressed as an inequality and solved for the allowable range of a design or process variable.
Inequalities also define domains. A square-root expression requires a nonnegative radicand, so the domain is a solution set to an inequality. A rational expression requires a nonzero denominator, which produces excluded values that appear as open circles or holes. These domain constraints are not optional bookkeeping; they determine which algebraic solutions are legitimate. This is why solving a radical inequality requires checking candidate regions against the original statement rather than trusting the algebra alone.
Systems and two-variable inequalities: finding the overlap
A one-variable inequality yields a region on a number line. A two-variable inequality such as y > 2x − 1 yields a half-plane bounded by the line y = 2x − 1. Graph the boundary line dashed for strict inequalities and solid for inclusive ones, then test one point not on the line. The origin (0, 0) is usually the easiest test: substitute and check whether the statement holds.
A system of inequalities is satisfied only where all regions overlap. That intersection is the feasible region, and its corner points are where optimization problems find their candidates. The logic carries over from number lines: with a system like x > 1 and x ≤ 5, the solution is the intersection (1, 5]. With contradictory constraints such as x > 5 and x < 2, the intersection is empty and the system has no solution.
The counterintuitive detail: solving a system by combining inequalities can produce a valid new inequality that is weaker than the original system. Adding x > 2 and y > 3 to conclude x + y > 5 is true, but x + y > 5 does not imply the originals. Combined inequalities are necessary consequences, not equivalent systems.
Comparison of solution methods
| Method | Best for | Boundary handling | Key failure mode |
|---|---|---|---|
| Algebraic isolation | Linear, simple polynomial | Reversal rule on ×/÷ by negative | Forgetting reversal |
| Sign analysis (test intervals) | Polynomial, rational | Critical points split line into intervals | Ignoring excluded zeros of denominator |
| Graphical half-plane test | Two-variable linear | Dashed vs solid boundary line | Shading without a test point |
| System intersection | Multiple constraints | Overlap region only | Assuming combined inequality is equivalent |
| Domain restriction check | Radical, rational | Excluded values as open points | Accepting extraneous solutions |
Timeline of a typical solution
| Step | Action | What to watch |
|---|---|---|
| 1 | Simplify each side | Distributing negatives correctly |
| 2 | Collect variable terms on one side | Reversal only on ×/÷ by negative |
| 3 | Isolate the variable | Keep inequality direction intact |
| 4 | State the boundary precisely | Open vs closed circle |
| 5 | Graph on number line or plane | Test point for two-variable cases |
| 6 | Express in interval notation | Parentheses around infinities |
| 7 | Verify with a sample value | Pick a point inside the shaded region |
Misconceptions and red flags
Treat every one of these as a warning sign in your own work or a student's:
Reversing the sign when adding or subtracting a negative number. Addition never reverses order; only multiplication and division by a negative do. Bracketing infinity, as in [3, ∞]. Shading without testing a point in two-variable problems, which frequently shades the wrong half-plane. Assuming a combined inequality is equivalent to its source system rather than merely implied by it. Multiplying both sides of an inequality by a variable expression without tracking where that expression is negative. Forgetting domain restrictions, then reporting a solution that makes the original expression undefined.
The variable-multiplication trap deserves emphasis because it appears in professional contexts. If you multiply an inequality by a quantity whose sign you have not established, you cannot know whether to reverse. The safe move is to split into cases or use sign analysis.
Practical actions and thresholds
Work through this sequence on any inequality problem, and the procedure becomes reliable:
Rewrite subtraction of negative terms as addition where useful, then simplify both sides completely before applying any operation. Move all variable terms to one side so the comparison reads variable versus constant. Isolate the variable, and pause at every multiplication or division to record the sign of the divisor. Reverse only if it is negative. Convert the result into both interval notation and inequality form, then compare them for consistency. For two-variable inequalities, graph the boundary as dashed (strict) or solid (inclusive), then substitute a test point. For systems, solve each inequality separately, graph all of them, and identify the intersection as the final solution set.
A useful verification threshold: pick one number inside your claimed solution and one outside, and check both against the original inequality. If either fails, the solution set or the boundary classification is wrong. This two-point check catches the majority of sign and shading errors.
Limitations and what this method cannot decide
The procedure above solves linear inequalities completely and polynomial and rational inequalities through sign analysis. It does not extend cleanly to inequalities involving transcendental functions, where test-interval methods become impractical without numerical tools. It also cannot resolve inequalities where the comparison operator is undefined, such as expressions involving division by zero at the boundary, which require separate treatment as excluded values.
Rounding introduces a subtle limitation in applied work. A threshold stated as "at least 12.5 units" differs operationally from one stated as "greater than 12.5 units," and converting between strict and inclusive forms changes which boundary values are admissible. In engineering and quality contexts, verify which convention the governing standard uses before graphing the boundary as open or closed, because the distinction determines whether a boundary-case product passes.