Math Made Practical — Percentages, Fractions & Statistics
School teaches procedures; life needs the meaning. This guide connects each core tool — percentages, fractions, statistics, equations — to the decisions where it actually earns its keep.
Most adult math fits in four toolboxes: percentages (change, discounts, rates), fractions (parts and ratios), descriptive statistics (averages and spread), and equations (solving for unknowns). Master these and budgets, grades, data, and deals become arithmetic instead of anxiety.
Each section explains the idea, the common trap, and links to a calculator that shows steps — built on exact standards (NIST constants, IEEE floating point) so the answers you get are the answers a mathematician would verify.
1. Percentages — The Multiplier Trap
Percent change always divides by the ORIGINAL value: 80 → 100 is +25%, but 100 → 80 is −20%. Gains and losses don't cancel — +50% then −50% ends at 75% of start. Successive changes multiply: 30% off plus 15% off is ×0.70 ×0.85 = ×0.595 (a 40.5% total discount, not 45%).
Percentage points ≠ percent: a rate moving 4% → 5% rose one point but 25% relatively. Finance headlines blur these constantly; the difference can be a decimal or a doubling.
2. Fractions — Arithmetic That Scales
Add/subtract via common denominators, multiply straight across, divide by flipping. The three skills that make it fast: simplifying by GCD (8/12 → 2/3), converting to mixed numbers for intuition (19/12 → 1 7/12), and cross-multiplying to compare (3/4 vs 5/7: 21 vs 20 → 3/4 wins).
The same arithmetic powers ratios in recipes, scale drawings, and medication dosing — the fraction calculator shows every step so the procedure sticks.
3. Descriptive Statistics — Mean vs Median vs Spread
Mean uses every value but chases outliers; median resists them (that's why home prices and salaries report medians). Standard deviation adds the spread story: ±1σ holds ~68% of bell-shaped data, making 'how unusual is this value' computable.
Sample vs population matters at the denominator: n−1 (sample) corrects the underestimation that using the sample mean causes. If you have ALL the data, divide by N; if it's a sample, n−1.
4. Solving Equations — Undo in Reverse Order
Linear equations solve by isolating x: undo add/subtract, then multiply/divide — reverse PEMDAS. Variables on both sides gather first. Always verify by substitution; the check catches the sign errors that cause 90% of wrong answers.
No-solution and infinite-solution cases are features: a false statement (5=3) means parallel lines; a true statement (5=5) means the same line. Recognizing them saves you from hunting an x that doesn't exist.
All Math Tools
Related Reading
Frequently Asked Questions
Why do +25% and −25% not cancel?
Different bases: up from 80 to 100 (+25%) then down 25% of 100 = 75. Recovering a −25% drop needs +33.3%. Percentages are multipliers, not reversible ticks.
When do I use median instead of mean?
Skewed or outlier-prone data: incomes, prices, response times. If mean − median exceeds ~10% of the mean, report the median — it describes the typical case the mean distorts.
Why divide by n−1 for sample standard deviation?
The sample mean absorbs one degree of freedom, underestimating spread if you divide by n. Bessel's correction (n−1) makes the sample estimate unbiased for the population.
What's the fastest way to compare fractions?
Cross-multiply: for a/b vs c/d, compare a×d vs c×b — no common denominator needed. It works because both sides scale to the same hidden denominator.
How do I check my algebra answer?
Substitute into the ORIGINAL equation. If both sides match, done. This 5-second habit catches sign and arithmetic slips that re-deriving misses.
Authoritative Sources
Last reviewed: September 2026.