Inequalities and Absolute Value: Protecting Signs Under Time Pressure
Most inequality errors are sign-handling errors; a small set of protective habits converts this topic from volatile to routine.
Why this matters
Inequalities punish habits imported from equation solving. Dividing both sides by a variable is safe in an equation and hazardous in an inequality, because a negative divisor silently reverses the direction of truth. Absolute value adds a second layer: one expression represents two cases, and dropping a case discards half the answer. These are not knowledge gaps so much as discipline gaps, which is why strong algebra students still miss them.
The exam leans on this topic in both Problem Solving and Data Sufficiency contexts, where the question is often whether a statement forces a sign or a range. Candidates who can manipulate inequalities without losing direction, and who split absolute values into cases automatically, gain accuracy exactly where the test expects carelessness.
A working method
Adopt three protective rules. First, never multiply or divide an inequality by any quantity whose sign is unknown; test cases or rearrange to avoid the operation. Second, when the sign is known and negative, reverse the inequality in the same stroke of the pen, not as an afterthought. Third, treat compound inequalities as a chain and apply every operation to all three parts simultaneously.
For absolute value, translate structurally before solving: an absolute value less than a constant means a bounded interval around the center, while greater than means two outward rays. For equations, split into the positive and negative cases and validate each candidate in the original statement, because case-splitting can manufacture extraneous solutions. When two absolute values interact, the number line, marking where each expression changes sign, is usually faster than blind case algebra.
How to practice this skill
Run a signed-numbers stress set: fifteen inequality problems in which at least one variable can be negative, zero, or a fraction between zero and one. Before solving, write the three test values you will try. Fractions between zero and one expose squaring assumptions, and zero exposes division hazards; making those test values a ritual converts trap-spotting into reflex.
Follow with an absolute value ladder: five interval translations, five two-case equations, and five number-line problems mixing two expressions. Time the second pass and log every miss as direction loss, dropped case, or unvalidated candidate. Each label has a distinct cure, and mixing them in review dilutes all three.
A rigorous review protocol
In review, reconstruct the exact step where direction or a case disappeared. Sign errors are almost never conceptual mysteries; they are single unguarded operations, and the log entry should quote the operation itself, such as divided by x without knowing its sign. Rewrite the step correctly and note the protective rule that would have caught it.
Then re-solve using a different representation: if you used algebra, redo it on a number line; if you used cases, redo it with test values. Agreement between two representations is strong evidence the answer is stable, and the comparison deepens the intuition that inequality statements describe regions, not single points.
Applying it in a timed section
Under time, prefer test values early on Data Sufficiency inequality statements: one negative, one zero when allowed, one fraction. Two well-chosen values often settle sufficiency faster than symbolic manipulation, and they are immune to direction-reversal slips. If algebra begins sprawling past a few lines, that sprawl is itself evidence you missed a structural translation, so step back once before pressing forward.
What mastery looks like
Mastery here looks like boring reliability: directions never flip unnoticed, cases never vanish, and fraction and zero test values appear without being summoned. When inequality items feel procedural instead of treacherous, you have converted the section's favorite carelessness trap into steady points.