Mathbench

NBT Mathematics (MAT) — Complete Study Guide

Read this first — it is written for you.

If maths has never quite "clicked" for you, this guide is built exactly for that. Nothing here assumes you already understand it. Every rule is explained in plain English before you see the formula, every symbol is spelled out in words, and every method comes with a fully worked example where every single step is shown — no steps are skipped and nothing is left for you to "just know."

You do not have to be naturally good at maths to do well on this test. You have to understand a small number of ideas properly, and practise them until they feel familiar. That is a skill anyone can build. Take it one small idea at a time, and be patient with yourself.

The National Benchmark Test in Mathematics (NBT MAT) checks whether you are ready for university-level maths. It is not the same as your matric exam. Matric often rewards long calculations. The NBT rewards understanding — knowing why a rule works, so you can use it on a question you have never seen before.

No calculator is allowed. That sounds scary, but it is actually good news: it means the numbers are almost always kept small and friendly on purpose. You will rarely need ugly arithmetic.


How This Book Is Organised

This opening guide is the map; the ten chapters after it are the territory. Here is how the whole book fits together:

How to work a question — the routine that makes this book work: attempt it honestly first, even if you only get partway. Then read the full explanation even when you got it right — the explanations teach the traps, shortcuts and cross-connections that turn one question into ten marks' worth of understanding. If you got it wrong, find your error among the "why the others are wrong" notes; recognising your own mistake pattern is the fastest improvement there is.

The five areas the NBT tests — and where to practise each

The NBT MAT draws every paper from the same five competence areas, named on the NBT's own website. This table is your route map:

NBT competence area Practise it in
Problem solving and modelling — algebraic processes (operations, ratios and percentages, surds, logarithms and exponents, financial calculations, algebraic manipulation) Chapters 1, 2, 7 and 8
Problem solving and modelling — functions represented by graphs and equations (domain and range, properties of graphs, transformations) Chapters 3 and 10
Basic trigonometry, including graphs, identities, equations, and the area, sine and cosine rules in 2D and 3D Chapter 4
Spatial perception — properties of 2D and 3D shapes, perimeter, area and volume, and analytic geometry Chapter 5
Data handling — measurement and representation, and the interpretation of both Chapter 6
Logical skills — making deductions and deciding whether a given assertion is true Every chapter; Chapter 11 leans on it hardest

How much of the paper is each one worth? Nobody outside the NBT knows. The NBT publishes the areas and the four cognitive levels, but it does not publish how many of the 60 questions fall in each area, and its National Report gives performance by subdomain rather than counts. Any book or website that prints an exact split is guessing — so this one does not. Practise all six rows; the test is only 60 questions and it mixes them.

(Chapter 10 covers introductory differential calculus, which the NBT's own reporting places under functions and graphs. Chapter 9's mock exams mix every area, in no claimed proportion.)

How to Use This Guide

The rest of this guide has five parts. You do not have to read them in order, and you definitely do not have to read them all in one sitting.

A promise about this test: there is no negative marking. You are never punished for a wrong answer. So you must answer every single question, even if you are guessing. A blank is a guaranteed zero; a guess has a 1-in-4 chance of being right. Never leave a bubble empty.

If you only remember five things from this whole guide: 1. Answer every question — a blank scores 0, a guess might score 1. 2. Learn the times tables, the special-angle table, and the basic formulas until you know them without thinking. 3. Turn "increase by 15%" into "multiply by 1.15" — percentages are just multiplying. 4. When stuck, try the four options one by one and see which fits. 5. Read the question slowly, twice. Half of all mistakes are just misreading.


The Language of Maths — Words Decoded

Most people who struggle with maths are not actually bad at maths — they get lost because of the words and symbols, which nobody ever explained in plain language. Here they are. Keep coming back to this list.

Everyday maths words

Word What it means in plain English
Sum The answer when you add. "The sum of 3 and 5" means 3+5=83 + 5 = 8.
Difference The answer when you subtract. "The difference of 8 and 3" means 83=58 - 3 = 5.
Product The answer when you multiply. "The product of 4 and 5" means 4×5=204 \times 5 = 20.
Quotient The answer when you divide. "The quotient of 20 and 4" means 20÷4=520 \div 4 = 5.
Term One "piece" of an expression, separated by ++ or - signs. In 3x+53x + 5, the terms are 3x3x and 55.
Expression A bit of maths with no equals sign, e.g. 3x+53x + 5. You can simplify it but not solve it.
Equation Two expressions with an equals sign between them, e.g. 3x+5=113x + 5 = 11. You solve it to find the unknown.
Variable A letter that stands for an unknown number, usually xx or yy.
Constant A plain number that does not change, like 55 or 2-2.
Coefficient The number multiplying a letter. In 3x3x, the coefficient is 33.
Substitute Replace a letter with a number. "Substitute x=2x = 2 into 3x3x" means work out 3×2=63 \times 2 = 6.
Simplify Make an expression shorter/neater without changing its value.
Solve Find the value of the unknown letter that makes an equation true.
Evaluate Work out the actual number an expression comes to.

Number words

Word What it means in plain English
Integer A whole number, positive or negative, including zero: ,2,1,0,1,2,\dots, -2, -1, 0, 1, 2, \dots (no fractions, no decimals).
Natural number The counting numbers: 1,2,3,4,1, 2, 3, 4, \dots (in South Africa these start at 1, not 0).
Rational number Any number you can write as a fraction, e.g. 34\tfrac{3}{4}, 0.50.5, 77, 0.3330.333\dots
Irrational number A number you cannot write as a neat fraction — its decimals go on forever with no pattern, e.g. 2\sqrt{2}, π\pi.
Factor A number that divides into another exactly. The factors of 1212 are 1,2,3,4,6,121, 2, 3, 4, 6, 12.
Multiple The times-table of a number. Multiples of 33 are 3,6,9,12,3, 6, 9, 12, \dots
Prime number A number with exactly two factors: 11 and itself (2,3,5,7,11,2, 3, 5, 7, 11, \dots). 11 is not prime.

Fraction and power words

Word What it means in plain English
Numerator The top of a fraction. In 34\tfrac{3}{4}, the numerator is 33.
Denominator The bottom of a fraction. In 34\tfrac{3}{4}, the denominator is 44.
Reciprocal A fraction flipped upside down. The reciprocal of 34\tfrac{3}{4} is 43\tfrac{4}{3}. Dividing by a fraction = multiplying by its reciprocal.
Exponent / power / index The small raised number telling you how many times to multiply. In 23=2×2×22^3 = 2 \times 2 \times 2, the exponent is 33.
Base The big number being multiplied. In 232^3, the base is 22.
Square Multiply a number by itself: 52=5×5=255^2 = 5 \times 5 = 25.
Square root (x\sqrt{\phantom{x}}) The reverse of squaring: "what number times itself gives this?" 25=5\sqrt{25} = 5.
Surd A square root that does not come out to a whole number, like 2\sqrt{2} or 5\sqrt{5}. We leave it as it is.

Symbols you will see

Symbol Read it as
== "is equal to"
\neq "is not equal to"
<< , >> "less than", "greater than"
\leq , \geq "less than or equal to", "greater than or equal to"
±\pm "plus or minus" (two answers: one with ++, one with -)
\therefore "therefore" (so...)
\approx "is approximately equal to"
\sum (sigma) "add up a list of things"
π\pi (pi) a fixed number 3.14\approx 3.14, used for circles
θ\theta (theta) a letter used to stand for an angle
x1,x2x_1, x_2 "x-one, x-two" — just a way to label a first and second value

A habit that will save you again and again: whenever a question confuses you, read it out loud in words, replacing every symbol with its meaning. Suddenly it is just a sentence you can understand.


The Test at a Glance: 60 Questions, 3 Hours, No Calculator

What you need to know Detail
How many questions 60, each multiple-choice with four options (A, B, C, D)
How long 3 hours — that is an average of 3 minutes per question
Calculator Not allowed
What you may take in Your ID, two sharpened pencils and an eraser, lunch and water
What you may not take in A calculator, a dictionary, scrap paper, notebooks, and any phone, tablet or smart watch
Wrong answers No penalty — so guess if you must, and never leave a blank
Question order All topics are mixed together — there are no sections

Is a calculator allowed in the NBT? No. The NBT's own rules list a calculator among the things you may not bring, along with a dictionary, notebooks and any phone, tablet or smart watch. That sounds harsh and is actually in your favour: it means the numbers are chosen to be workable by hand, and Part 2 of this guide is nothing but ways to do that quickly.

And there is no scrap paper either. You work in the white space of the question booklet, so get used to laying out your working small and neat while you practise. Bathroom breaks are not allowed during the session, so plan for three hours in your seat.

What this means for you: you have plenty of time (3 minutes each is generous), so do not rush. Do the easy ones first to bank those marks, then come back for the hard ones.


Part 1 — The Building Blocks: Your NBT Formula Sheet

This is the actual maths. Taken together, the tables below are a formula sheet for the NBT — every formula, rule and fact this test draws on, gathered in one place so that you can learn them and check them. You walk into the test with nothing but pencils, so the sheet has to be in your head; that is exactly what these tables are for.

Each topic below follows the same friendly pattern:

You do not need to master everything at once. Pick one topic, understand it, practise it, then move on. Little by little.


1. Number Sense

What this is really about: being comfortable with fractions, decimals, percentages, and different kinds of numbers. This is the most common topic on the test, and the good news is it is the most learnable. Get this solid and you have already earned a big chunk of the marks.

1.1 Fraction–Decimal–Percentage Table

A fraction, a decimal, and a percentage are three different ways of writing the same amount. 12\tfrac{1}{2}, 0.50.5, and 50%50\% are all "half." Learn this table by heart — it saves you from doing conversions under pressure.

Fraction Decimal Percentage
12\dfrac{1}{2} 0.50.5 50%50\%
14\dfrac{1}{4} 0.250.25 25%25\%
34\dfrac{3}{4} 0.750.75 75%75\%
15\dfrac{1}{5} 0.20.2 20%20\%
25\dfrac{2}{5} 0.40.4 40%40\%
35\dfrac{3}{5} 0.60.6 60%60\%
45\dfrac{4}{5} 0.80.8 80%80\%
18\dfrac{1}{8} 0.1250.125 12.5%12.5\%
38\dfrac{3}{8} 0.3750.375 37.5%37.5\%
58\dfrac{5}{8} 0.6250.625 62.5%62.5\%
78\dfrac{7}{8} 0.8750.875 87.5%87.5\%
110\dfrac{1}{10} 0.10.1 10%10\%
13\dfrac{1}{3} 0.30.\overline{3} 33.3%33.\overline{3}\%
23\dfrac{2}{3} 0.60.\overline{6} 66.6%66.\overline{6}\%

(The little bar in 0.30.\overline{3} means the 3 repeats forever: 0.33330.3333\dots)

To turn a fraction into a percentage: divide top by bottom, then multiply by 100. To turn a percentage into a fraction: put it over 100 and simplify.

1.2 Number Facts to Know Instantly

You will use these constantly. Flash-card them until they are automatic.

Perfect squares (a number times itself), from 121^2 to 15215^2:

1, 4, 9, 16, 25, 36, 49, 64, 81, 100, 121, 144, 169, 196, 2251,\ 4,\ 9,\ 16,\ 25,\ 36,\ 49,\ 64,\ 81,\ 100,\ 121,\ 144,\ 169,\ 196,\ 225

Perfect cubes (a number times itself, times itself again), 131^3 to 535^3:

1, 8, 27, 64, 1251,\ 8,\ 27,\ 64,\ 125

Powers of 2 (keep doubling), up to 2102^{10}:

2, 4, 8, 16, 32, 64, 128, 256, 512, 10242,\ 4,\ 8,\ 16,\ 32,\ 64,\ 128,\ 256,\ 512,\ 1024

Prime numbers up to 50 (only divisible by 1 and themselves):

2, 3, 5, 7, 11, 13, 17, 19, 23, 29, 31, 37, 41, 43, 472,\ 3,\ 5,\ 7,\ 11,\ 13,\ 17,\ 19,\ 23,\ 29,\ 31,\ 37,\ 41,\ 43,\ 47

The 10% ladder — powers of 1.11.1, which the NBT uses constantly in interest and growth questions (notice the digits are the rows of Pascal's triangle):

1.12=1.211.13=1.3311.14=1.46411.1^2 = 1.21 \qquad 1.1^3 = 1.331 \qquad 1.1^4 = 1.4641

Triangular numbers — the running totals 1+2+3++n1+2+3+\cdots+n (formula n(n+1)2\tfrac{n(n+1)}{2}). Know at least: n=515n=5 \to 15,  n=1055\ n=10 \to 55,  n=20210\ n=20 \to 210,  n=40820\ n=40 \to 820.

Sum of the first nn odd numbers is always n2n^2: 1+3+5+7=16=421+3+5+7 = 16 = 4^2. This one fact is a confirmed exam favourite.

Parity in one breath (odd/even rules): odd ± odd = even; even ± even = even; odd ± even = odd; odd × odd = odd; even × anything = even.

1.3 Percentages — the one idea that unlocks everything

In plain English: "per cent" literally means "out of 100." So 30%30\% just means 30100=0.3\tfrac{30}{100} = 0.3. The single most useful trick in the whole test is this:

Any percentage change is just a multiplication. To increase by 15%15\%, multiply by 1.151.15. To decrease by 15%15\%, multiply by 0.850.85.

Why? If you have R100 and add 15%, you now have the original 100% plus 15% =115%= 115\% of what you had =1.15×100=115= 1.15 \times 100 = 115. To take away 15%, you keep 100%15%=85%=0.85100\% - 15\% = 85\% = 0.85 of it.

What you want to do How to do it
Find p%p\% of a number nn p100×n\dfrac{p}{100} \times n
Percentage increase NewOldOld×100\dfrac{\text{New} - \text{Old}}{\text{Old}} \times 100
Percentage decrease OldNewOld×100\dfrac{\text{Old} - \text{New}}{\text{Old}} \times 100
Increase a number by p%p\% Multiply by (1+p100)\left(1 + \dfrac{p}{100}\right)
Decrease a number by p%p\% Multiply by (1p100)\left(1 - \dfrac{p}{100}\right)
Find the original before a change Divide by the multiplier (don't add the percentage back!)

Worked example — a shirt in a sale: A shirt costs R240. It is increased by 15%. What is the new price?

Watch out — the "back to the start" trap: A price goes down 10%, then up 10%. You are not back where you started. Down 10% means ×0.9\times 0.9; up 10% means ×1.1\times 1.1. Together: 0.9×1.1=0.990.9 \times 1.1 = 0.99, which is a 1% loss. Whenever percentages happen one after another, multiply the multipliers — never add or subtract the percentages.

1.4 Types of Numbers

Numbers live in nested groups, like boxes inside boxes:

NZQR\mathbb{N} \subset \mathbb{Z} \subset \mathbb{Q} \subset \mathbb{R}

Read left to right, that says: natural numbers are inside the integers, which are inside the rationals, which are inside the reals.

Quick test for a square root: n\sqrt{n} is a nice rational number only if nn is a perfect square. 25=5\sqrt{25} = 5 (rational), but 5\sqrt{5} is irrational.


1.5 Scientific Notation

Scientific notation writes a number as a×10na \times 10^n, where aa is between 11 and 1010 and nn is a whole number. The exponent counts the places the decimal point moves — positive for big numbers, negative for small ones.

4500000=4.5×1060.00045=4.5×1044\,500\,000 = 4.5 \times 10^6 \qquad\qquad 0.00045 = 4.5 \times 10^{-4}

To multiply or divide, deal with the coefficients and the powers of ten separately:

If the coefficient lands outside 11 to 1010, fix it: 37×104=3.7×10537 \times 10^4 = 3.7 \times 10^5.

1.6 Speed, Distance and Time

One relationship, read three ways:

distance=speed×timespeed=distancetimetime=distancespeed\text{distance} = \text{speed} \times \text{time} \qquad \text{speed} = \frac{\text{distance}}{\text{time}} \qquad \text{time} = \frac{\text{distance}}{\text{speed}}

Two rules stop nearly every mistake here:

  1. Make the units agree first. With a speed in km/h, 3030 minutes is 0.50.5 hours — divide minutes by 6060, never treat them as decimals.
  2. Average speed is total distance ÷ total time. It is not the average of the speeds. If you travel equal distances at two different speeds, more time is spent at the slower one, so the true average always comes out below the halfway point between them.

2. Exponents and Logarithms

What this is really about: an exponent (or "power") is just a shortcut for repeated multiplying. 242^4 means "multiply four 2's together": 2×2×2×2=162 \times 2 \times 2 \times 2 = 16. A logarithm is the reverse question: "how many 2's did I multiply to get 16?" — the answer is 4. That is all a log is: an exponent, asked backwards.

2.1 Laws of Exponents

These rules let you combine powers quickly. The trick to all of them: they only work when the base (the big number) is the same.

Law Formula In plain English
Multiplying aman=am+na^m \cdot a^n = a^{m+n} Same base, multiplying → add the powers
Dividing aman=amn\dfrac{a^m}{a^n} = a^{m-n} Same base, dividing → subtract the powers
Power of a power (am)n=amn(a^m)^n = a^{mn} A power raised to a power → multiply the powers
Negative power an=1ana^{-n} = \dfrac{1}{a^n} A minus sign in the power means "flip it to the bottom"
Zero power a0=1a^0 = 1 Anything (except 0) to the power 0 is just 11
Fractional power amn=amna^{\frac{m}{n}} = \sqrt[n]{a^m} Bottom of the fraction = the root; top = the power
Power of a product (ab)n=anbn(ab)^n = a^n b^n Share the power out to each factor

Worked example: Simplify x5x3x4\dfrac{x^5 \cdot x^3}{x^4}.

2.2 Laws of Logarithms

A logarithm answers "what power?" Written logaN=x\log_a N = x, it means "aa to the power xx gives NN," i.e. ax=Na^x = N. The two are the same fact written two ways:

y=ax    x=logayy = a^x \iff x = \log_a y

Law Formula In plain English
Product loga(xy)=logax+logay\log_a(xy) = \log_a x + \log_a y Log of a productadd the logs
Quotient loga ⁣(xy)=logaxlogay\log_a\!\left(\dfrac{x}{y}\right) = \log_a x - \log_a y Log of a divisionsubtract the logs
Power loga(xn)=nlogax\log_a(x^n) = n\log_a x A power inside → comes out to the front as a multiplier
Log of the base logaa=1\log_a a = 1 "What power of aa gives aa?" → 11
Log of 1 loga1=0\log_a 1 = 0 "What power gives 1?" → 00

Worked example: Evaluate log232\log_2 32.

Watch out — the biggest log/exponent trap: you can only split a log when things are multiplied or divided inside. log(a+b)\log(a+b) does NOT equal loga+logb\log a + \log b. There is no rule for adding inside a log. (The same idea catches people with roots: a+ba+b\sqrt{a+b} \neq \sqrt{a} + \sqrt{b}.)


3. Algebra and Surds

What this is really about: algebra is arithmetic with letters standing in for numbers. "Factorising" means writing an expression as a multiplication (breaking it into pieces multiplied together); "expanding" is the reverse (multiplying the brackets out).

3.1 Special Identities — patterns worth memorising

These come up so often that recognising them instantly saves huge time.

Pattern Formula
Square of a sum (a+b)2=a2+2ab+b2(a + b)^2 = a^2 + 2ab + b^2
Square of a difference (ab)2=a22ab+b2(a - b)^2 = a^2 - 2ab + b^2
Difference of two squares a2b2=(ab)(a+b)a^2 - b^2 = (a - b)(a + b)
Sum of cubes a3+b3=(a+b)(a2ab+b2)a^3 + b^3 = (a + b)(a^2 - ab + b^2)
Difference of cubes a3b3=(ab)(a2+ab+b2)a^3 - b^3 = (a - b)(a^2 + ab + b^2)

The one everyone forgets: (a+b)2(a+b)^2 is not a2+b2a^2 + b^2. You must include the middle term 2ab2ab. Check with numbers: (2+3)2=25(2+3)^2 = 25, but 22+32=132^2 + 3^2 = 13. Different! The missing piece is the 2ab=2×2×3=122ab = 2\times2\times3 = 12, and 13+12=2513 + 12 = 25. ✓

3.2 Factorising — always try these in order

When asked to factorise, run down this checklist from the top:

  1. Common factor first — is there a number or letter in every term you can pull out? Always check this first. 6x+9=3(2x+3)6x + 9 = 3(2x + 3).
  2. Difference of two squares — does it look like a2b2a^2 - b^2? Then =(ab)(a+b)= (a-b)(a+b).
  3. Sum or difference of cubesa3±b3a^3 \pm b^3.
  4. Trinomial (three terms, x2+bx+cx^2 + bx + c) — find two numbers that multiply to cc and add to bb.
  5. Grouping — for four terms, group them in pairs.
  6. Factor theorem — for cubics: if putting x=px = p into the expression gives 00, then (xp)(x - p) is a factor.

Worked example — a trinomial: Factorise x2+5x+6x^2 + 5x + 6.

3.3 Surds (leftover square roots)

A surd is a root that does not simplify to a whole number, like 2\sqrt{2}. Rules:

ab=abab=ab\sqrt{ab} = \sqrt{a}\cdot\sqrt{b} \qquad \sqrt{\frac{a}{b}} = \frac{\sqrt{a}}{\sqrt{b}}

Simplifying a surd — pull out the biggest perfect-square factor hiding inside:

72=36×2=36×2=62\sqrt{72} = \sqrt{36 \times 2} = \sqrt{36}\times\sqrt{2} = 6\sqrt{2}

Adding surds — you can only add "like" surds (same number under the root), exactly like adding 2x+3x=5x2x + 3x = 5x:

12+27=23+33=53\sqrt{12} + \sqrt{27} = 2\sqrt{3} + 3\sqrt{3} = 5\sqrt{3}

Rationalising the denominator — getting rid of a root on the bottom of a fraction. Multiply top and bottom by the conjugate (the same expression with the middle sign flipped):

351×5+15+1=3(5+1)51=3(5+1)4\frac{3}{\sqrt{5} - 1} \times \frac{\sqrt{5} + 1}{\sqrt{5} + 1} = \frac{3(\sqrt{5} + 1)}{5 - 1} = \frac{3(\sqrt{5} + 1)}{4}

3.4 Completing the Square

This rewrites x2+bx+cx^2 + bx + c into the form (x+p)2+q(x + p)^2 + q, which instantly shows the turning point of a parabola. The recipe:

Worked example: Write x2+8x+3x^2 + 8x + 3 as (x+p)2+q(x+p)^2 + q.


4. Quadratic Equations

What this is really about: a "quadratic" is any equation with an x2x^2 in it (and no higher power). Its graph is a U-shape (or upside-down U) called a parabola. Solving it means finding where it crosses the x-axis.

4.1 Three Ways to Solve

  1. Factorising — write it as (xp)(xq)=0(x - p)(x - q) = 0. Because anything times zero is zero, either xp=0x - p = 0 or xq=0x - q = 0, so x=px = p or x=qx = q. Always try this first.
  2. The quadratic formula — works for every quadratic ax2+bx+c=0ax^2 + bx + c = 0, even ugly ones:

x=b±b24ac2a\boxed{\,x = \dfrac{-b \pm \sqrt{b^2 - 4ac}}{2a}\,}

  1. Completing the square — sometimes handy, but the formula usually wins.

Worked example — by factorising: Solve x27x+10=0x^2 - 7x + 10 = 0.

4.2 The Discriminant — a shortcut to the "nature of the roots"

Sometimes a question does not want you to solve, only to say how many real answers there are. The part under the root in the formula, Δ=b24ac\Delta = b^2 - 4ac (called the discriminant), tells you without doing the whole calculation:

If Δ=b24ac\Delta = b^2 - 4ac is... Then the equation has... The parabola...
positive and a perfect square two different rational answers crosses the x-axis at two nice points
positive, not a perfect square two different irrational answers crosses the x-axis at two points
zero one repeated answer just touches the x-axis
negative no real answers never touches the x-axis

(A negative under a square root has no real answer — that is your signal for "no real roots.")

4.3 Sum and Product of the Roots

For ax2+bx+c=0ax^2 + bx + c = 0 with answers x1x_1 and x2x_2, without solving you already know:

x1+x2=bax1x2=cax_1 + x_2 = -\frac{b}{a} \qquad\qquad x_1 \cdot x_2 = \frac{c}{a}

These let you check answers or solve clever questions without ever finding x1x_1 and x2x_2 separately.

4.4 Quadratic Inequalities — solve with a picture

To solve something like x2x6>0x^2 - x - 6 > 0: factorise, find the two roots, then think of the parabola. An upward parabola is below zero between its roots and above zero outside them.

Watch out: before any of this, ask whether the statement can even be false — x2+4<0x^2 + 4 < 0 has no solutions (a square is never negative), and the NBT loves testing exactly that.


5. Functions and Graphs

What this is really about: a function is a rule that takes an input number and gives back one output number — like a machine. f(x)=2x+1f(x) = 2x + 1 means "double it and add one." The graph is just a picture of every input-output pair plotted as points.

Two bits of language:

5.1 The Main Graph Types

Function Standard equation What its graph looks like
Linear (straight line) y=mx+cy = mx + c A straight line. mm = gradient (steepness), cc = where it crosses the y-axis
Quadratic (parabola) y=a(xp)2+qy = a(x - p)^2 + q A U-shape. Turning point at (p,q)(p,\,q). Opens up if a>0a>0, down if a<0a<0
Exponential y=abx+qy = a \cdot b^x + q A curve that shoots up (or decays down). Flat asymptote at y=qy = q
Hyperbola y=axp+qy = \dfrac{a}{x - p} + q Two curved branches. Asymptotes at x=px = p and y=qy = q
Logarithm y=logbxy = \log_b x The mirror-image of the exponential curve
Absolute value y=axp+qy = a\lvert x - p\rvert + q A V-shape. Vertex at (p,q)(p,\,q). Opens up if a>0a>0, down if a<0a<0

For a parabola y=ax2+bx+cy = ax^2 + bx + c, the turning point sits on the line

x=b2a.x = -\frac{b}{2a}.

Find that xx, then substitute it back in to get the yy of the turning point.

Worked example — turning point: Find the turning point of y=x26x+5y = x^2 - 6x + 5.

Why asymptotes happen (plain English): for a hyperbola, you can never divide by zero — so at the xx that makes the bottom zero, the graph "breaks" (vertical asymptote). And as xx gets huge, the fraction shrinks to almost nothing, so yy flattens out towards qq (horizontal asymptote).

5.2 Moving Graphs Around (Transformations)

These work for any graph y=f(x)y = f(x). Learn the pattern once and it applies everywhere.

What you do New equation Effect
Add kk on the outside y=f(x)+ky = f(x) + k Whole graph moves up kk
Subtract kk on the outside y=f(x)ky = f(x) - k Whole graph moves down kk
h-h inside the bracket y=f(xh)y = f(x - h) Moves right hh (yes, minus means right — this surprises everyone)
+h+h inside the bracket y=f(x+h)y = f(x + h) Moves left hh
Minus on the outside y=f(x)y = -f(x) Flips upside-down (reflects in the x-axis)
Minus on the inside y=f(x)y = f(-x) Flips left-right (reflects in the y-axis)

5.3 Inverse Functions

The inverse undoes what the function did. To find it: swap xx and yy, then solve for yy. On a graph, the inverse is the original reflected in the line y=xy = x. Handy fact: the inverse of y=bxy = b^x is y=logbxy = \log_b x.


6. Trigonometry

What this is really about: trigonometry ("trig") connects the angles of a right-angled triangle to the lengths of its sides. It sounds intimidating but rests on one memory trick: SOHCAHTOA.

6.1 SOHCAHTOA — the whole foundation

In a right-angled triangle, name the three sides from the angle you care about:

sinθ=OppositeHypotenusecosθ=AdjacentHypotenusetanθ=OppositeAdjacent\sin\theta = \frac{\text{Opposite}}{\text{Hypotenuse}} \qquad \cos\theta = \frac{\text{Adjacent}}{\text{Hypotenuse}} \qquad \tan\theta = \frac{\text{Opposite}}{\text{Adjacent}}

The nonsense word SOH-CAH-TOA stores all three: Sin = Opp/Hyp, Cos = Adj/Hyp, Tan = Opp/Adj.

Worked example: A right triangle has an opposite side of 3 and a hypotenuse of 5. Find sinθ\sin\theta.

6.2 Special Angles — memorise this table

For a few "nice" angles, the answers are exact. These appear constantly (remember, no calculator):

θ\theta sinθ\sin\theta cosθ\cos\theta tanθ\tan\theta
0° 00 11 00
30°30° 12\dfrac{1}{2} 32\dfrac{\sqrt{3}}{2} 13\dfrac{1}{\sqrt{3}}
45°45° 22\dfrac{\sqrt{2}}{2} 22\dfrac{\sqrt{2}}{2} 11
60°60° 32\dfrac{\sqrt{3}}{2} 12\dfrac{1}{2} 3\sqrt{3}
90°90° 11 00 undefined

6.3 The CAST Diagram — which sign in which quadrant

Angles bigger than 90° live in different "quadrants" (quarters) of a circle. In each quarter, some of sin/cos/tan are positive and some negative. CAST tells you which are positive:

Quadrant Angles Positive there Letter
I 0°90°90° all three All
II 90°90°180°180° only sin Sin
III 180°180°270°270° only tan Tan
IV 270°270°360°360° only cos Cos

Read the letters going round: A, S, T, C — remember it as "All Students Take Calculus."

6.4 The Identities You Must Know

An identity is a trig equation that is always true. The most important one:

sin2θ+cos2θ=1tanθ=sinθcosθ\sin^2\theta + \cos^2\theta = 1 \qquad\qquad \tan\theta = \frac{\sin\theta}{\cos\theta}

(The notation sin2θ\sin^2\theta means (sinθ)2(\sin\theta)^2 — square the whole thing.) From these you can also get 1+tan2θ=sec2θ1 + \tan^2\theta = \sec^2\theta.

The double-angle formulas — confirmed favourites on recent papers (practise them in Chapter 4):

sin2θ=2sinθcosθcos2θ=12sin2θ=2cos2θ1=cos2θsin2θ\sin 2\theta = 2\sin\theta\cos\theta \qquad\qquad \cos 2\theta = 1 - 2\sin^2\theta = 2\cos^2\theta - 1 = \cos^2\theta - \sin^2\theta

How they appear in disguise: an expression like 24sin23x2 - 4\sin^2 3x is really 2(12sin23x)=2cos6x2(1 - 2\sin^2 3x) = 2\cos 6x — factor until the bracket matches the identity, then remember the formula doubles whatever angle is inside (3x3x becomes 6x6x).

6.5 Reduction Formulae (angles outside 0–90°)

These convert a big or awkward angle back to a friendly one:

Angle sin\sin cos\cos tan\tan
180°θ180° - \theta +sinθ+\sin\theta cosθ-\cos\theta tanθ-\tan\theta
180°+θ180° + \theta sinθ-\sin\theta cosθ-\cos\theta +tanθ+\tan\theta
360°θ360° - \theta sinθ-\sin\theta +cosθ+\cos\theta tanθ-\tan\theta
90°θ90° - \theta cosθ\cos\theta sinθ\sin\theta
θ-\theta sinθ-\sin\theta +cosθ+\cos\theta tanθ-\tan\theta

How to use them: find the "reference angle" (the small acute angle to the horizontal axis), then use CAST to decide whether the answer is ++ or -.

6.6 Rules for Any Triangle (not just right-angled)

When a triangle has no right angle, use these:

Sine rule:asinA=bsinB=csinC\text{Sine rule:}\quad \frac{a}{\sin A} = \frac{b}{\sin B} = \frac{c}{\sin C}

Cosine rule:c2=a2+b22abcosC\text{Cosine rule:}\quad c^2 = a^2 + b^2 - 2ab\cos C

Area of a triangle:Area=12absinC\text{Area of a triangle:}\quad \text{Area} = \tfrac{1}{2}\,ab\sin C

(In these, a small letter is the side opposite the matching capital-letter angle.) Use the cosine rule when you know two sides and the angle between them, or all three sides. Use the sine rule for most other cases.


7. Analytical Geometry

What this is really about: doing geometry using coordinates — points written as (x,y)(x, y) on a grid. Every question here is really just Pythagoras or straight lines in disguise.

I want to find... Formula In plain English
Distance between two points d=(x2x1)2+(y2y1)2d = \sqrt{(x_2 - x_1)^2 + (y_2 - y_1)^2} Pythagoras: square the across-gap, square the up-gap, add, square-root
Midpoint (the middle) M=(x1+x22, y1+y22)M = \left(\dfrac{x_1 + x_2}{2},\ \dfrac{y_1 + y_2}{2}\right) The average of the two x's, and of the two y's
Gradient (steepness) m=y2y1x2x1m = \dfrac{y_2 - y_1}{x_2 - x_1} "Rise over run" — up-change divided by across-change
Equation of a line (know gradient + a point) yy1=m(xx1)y - y_1 = m(x - x_1) Slot in the gradient and the point
Parallel lines m1=m2m_1 = m_2 Same steepness
Perpendicular lines (meet at 90°) m1×m2=1m_1 \times m_2 = -1 Flip one gradient upside-down and change its sign
Circle centred at the origin x2+y2=r2x^2 + y^2 = r^2 rr is the radius
Circle centred at (a,b)(a, b) (xa)2+(yb)2=r2(x - a)^2 + (y - b)^2 = r^2

Worked example — distance: Find the distance between (1,2)(1, 2) and (4,6)(4, 6).

Watch out: for a perpendicular gradient you need both changes — flip it over and swap the sign. The perpendicular of 23\tfrac{2}{3} is 32-\tfrac{3}{2}, not just 23-\tfrac{2}{3} or 32\tfrac{3}{2}.


8. Geometry and Measurement

What this is really about: shapes, and how to find their area (the space inside), perimeter (the distance around), and for 3-D solids, volume (how much fits inside) and surface area (the total skin).

8.1 Flat Shapes — Area and Perimeter

Shape Area Perimeter
Square s2s^2 4s4s
Rectangle ×b\ell \times b 2(+b)2(\ell + b)
Triangle 12bh\dfrac{1}{2} b h add up the 3 sides
Circle πr2\pi r^2 2πr2\pi r (the "perimeter" of a circle is called the circumference)
Parallelogram bhb h 2(a+b)2(a + b)
Trapezium 12(a+b)h\dfrac{1}{2}(a + b)h add up all 4 sides

(Here hh always means the perpendicular height — straight up, at a right angle to the base — never a slanted side.)

8.2 Solid Shapes — Volume and Surface Area

Solid Volume Surface Area
Rectangular box ×b×h\ell \times b \times h 2(b+h+bh)2(\ell b + \ell h + bh)
Cylinder (tin can) πr2h\pi r^2 h 2πr2+2πrh2\pi r^2 + 2\pi r h
Sphere (ball) 43πr3\dfrac{4}{3}\pi r^3 4πr24\pi r^2
Cone 13πr2h\dfrac{1}{3}\pi r^2 h πr2+πr\pi r^2 + \pi r \ell
Pyramid 13×(base area)×h\dfrac{1}{3} \times (\text{base area}) \times h base + the slanted faces

(For the cone, \ell is the slant height — the sloped distance up the side.) Notice cones and pyramids both have a 13\tfrac{1}{3} — they hold exactly one-third of the box or cylinder they fit inside.

Worked example — cylinder volume: A tin can has radius 3 cm and height 10 cm. Find its volume (leave π\pi in the answer).

8.3 Theorems Worth Knowing

Pythagoras' theorem (right-angled triangles only): the two shorter sides squared, added, equal the longest side squared:

a2+b2=c2(c is the hypotenuse, the longest side)a^2 + b^2 = c^2 \quad (c \text{ is the hypotenuse, the longest side})

Handy triples to recognise instantly: 3-4-5, 5-12-13, 8-15-17, 7-24-25 (and any multiple, like 6-8-10).

Similar shapes and scale factor: if two shapes are the same shape but different sizes, and lengths are scaled by kk, then areas scale by k2k^2 and volumes by k3k^3. (Double the length → 4× the area → 8× the volume.)

Circle theorems to recognise:


9. Sequences and Series

What this is really about: a sequence is a list of numbers following a pattern (like 2,5,8,11,2, 5, 8, 11, \dots). A series is what you get when you add the numbers of a sequence up. There are two patterns to know.

9.1 Arithmetic — adding the same amount each time

An arithmetic sequence goes up (or down) by a fixed step, called the common difference dd. Example: 5,8,11,14,5, 8, 11, 14, \dots has d=3d = 3.

Tn=a+(n1)dSn=n2[2a+(n1)d]T_n = a + (n-1)d \qquad\qquad S_n = \frac{n}{2}\bigl[2a + (n-1)d\bigr]

Worked example: Find the 10th term of 5,8,11,14,5, 8, 11, 14, \dots

9.2 Geometric — multiplying by the same amount each time

A geometric sequence multiplies by a fixed number, the common ratio rr. Example: 3,6,12,24,3, 6, 12, 24, \dots has r=2r = 2.

Tn=arn1Sn=a(rn1)r1T_n = a\,r^{\,n-1} \qquad\qquad S_n = \frac{a(r^n - 1)}{r - 1}

Worked example: Find the 5th term of 2,6,18,2, 6, 18, \dots

Sum to infinity — if the numbers are shrinking (that is, r<1|r| < 1), an endless sum can still reach a finite total:

S=a1r(only works when r<1)S_\infty = \frac{a}{1 - r} \qquad (\text{only works when } |r| < 1)

9.3 Sigma Notation (the \sum symbol)

\sum (the Greek letter "sigma") just says "add up." i=15i\displaystyle\sum_{i=1}^{5} i means "add the whole numbers from 1 to 5" =1+2+3+4+5=15= 1+2+3+4+5 = 15. Useful shortcuts:

i=1ni=n(n+1)2i=1n(2i1)=n2i=1nc=cn\sum_{i=1}^{n} i = \frac{n(n+1)}{2} \qquad \sum_{i=1}^{n} (2i-1) = n^2 \qquad \sum_{i=1}^{n} c = c\,n

(That middle one says: the sum of the first nn odd numbers is exactly n2n^2 — the same fact from the number-facts list, now in sigma costume. The exam bank uses it in exactly this costume.)


10. Financial Mathematics

What this is really about: money growing (interest on savings) or shrinking (depreciation — things losing value). It is really just percentages applied over several years.

Situation Formula The idea
Simple interest A=P(1+in)A = P(1 + i n) Same fixed amount added every year
Compound growth A=P(1+i)nA = P(1 + i)^n Interest earns its own interest — grows faster
Straight-line depreciation A=P(1in)A = P(1 - i n) Value drops by the same rand amount each year
Depreciation (reducing balance) A=P(1i)nA = P(1 - i)^n Value shrinks by a % of its current worth each year
Inflation (future prices) A=P(1+i)nA = P(1 + i)^n Prices compound exactly like interest
Compounded mm times a year A=P ⁣(1+im)mnA = P\!\left(1 + \dfrac{i}{m}\right)^{mn} Split the year into mm pieces

(PP = starting amount, ii = interest rate as a decimal — so 8% becomes 0.08 — nn = number of years.)

Three more money situations the NBT tests (all practised in Chapter 7):

Simple vs compound, in one line: simple interest grows in a straight line; compound interest snowballs (each year's interest is bigger than the last). Over many years, compound is much larger.

Worked example — compound growth: R10 000 is invested at 10% per year, compounded, for 2 years.


11. Data Handling and Probability

What this is really about: two things — statistics (describing a set of numbers, like an average) and probability (measuring how likely something is).

11.1 Averages and Spread

Measure What it is
Mean (xˉ)(\bar{x}) The ordinary average: add everything up, divide by how many. Gets pulled around by very big/small values.
Median The middle number once the data is sorted in order. Barely affected by outliers.
Mode The value that appears most often.
Range Biggest minus smallest (maxmin\text{max} - \text{min}).
IQR (interquartile range) Q3Q1Q_3 - Q_1 — the spread of the middle half of the data.

The one rule people forget: to find the median, you must sort the numbers first. The median of 7,2,97, 2, 9 is not 22 — sort them to 2,7,92, 7, 9 and the middle one is 77.

Worked example — mean: Find the mean of 60,72,68,75,8060, 72, 68, 75, 80.

11.2 Probability

Probability is always a number between 0 (impossible) and 1 (certain):

P(event)=number of ways it can happentotal number of equally likely outcomesP(\text{event}) = \frac{\text{number of ways it can happen}}{\text{total number of equally likely outcomes}}

Rule Formula Plain English
Complement (the "opposite") P(not A)=1P(A)P(\text{not } A) = 1 - P(A) Chance it doesn't happen = 1 minus chance it does
Either A or B P(AB)=P(A)+P(B)P(AB)P(A \cup B) = P(A) + P(B) - P(A \cap B) Add them, but subtract the overlap so you don't count it twice
Both A and B (independent) P(AB)=P(A)×P(B)P(A \cap B) = P(A) \times P(B) If they don't affect each other, multiply

Worked example: A fair die is rolled. What is the probability of an even number?

11.3 Counting

Fundamental counting principle: if one choice can be made in mm ways and the next in nn ways, together there are m×nm \times n ways. (3 shirts and 4 trousers give 3×4=123\times 4 = 12 outfits.)

(The symbol n!n! — "n factorial" — means multiply all whole numbers from nn down to 1. So 4!=4×3×2×1=244! = 4\times3\times2\times1 = 24. And 0!=10! = 1.)

Worked example — does order matter? From 5 students, how many ways can you choose (a) a president and a secretary, (b) a pair of representatives?


Part 2 — Doing Maths in Your Head

With no calculator, how you calculate matters as much as what you calculate. These tricks keep the numbers small and friendly.

Turn decimals into fractions

Fractions are usually faster and safer by hand:

0.375×64=38×64=3×8=240.375 \times 64 = \frac{3}{8} \times 64 = 3 \times 8 = 24

Trap a square root between two you know

You do not need the exact value — just enough to pick the right option:

64<85<100    9<85<10\sqrt{64} < \sqrt{85} < \sqrt{100} \implies 9 < \sqrt{85} < 10

Since 85 is close to 81=9281 = 9^2, you can say 859.2\sqrt{85} \approx 9.2. Often that is enough to knock out the wrong options.

Break big roots into pieces

1296=36×36=36720=144×5=125\sqrt{1296} = \sqrt{36 \times 36} = 36 \qquad \sqrt{720} = \sqrt{144 \times 5} = 12\sqrt{5}

Cancel before you multiply

Cancelling first keeps the numbers tiny:

38×169=3×168×9=1×21×3=23\frac{3}{8} \times \frac{16}{9} = \frac{3 \times 16}{8 \times 9} = \frac{1 \times 2}{1 \times 3} = \frac{2}{3}

(The 3 cancels into the 9, and the 8 cancels into the 16, before you multiply.)

Percentages are just multipliers

Increase by 15% → ×1.15\times 1.15. Decrease by 15% → ×0.85\times 0.85. That one habit turns most "money" questions into a single multiplication.


Part 3 — Exam Strategy

Knowing the maths is only half the battle. These are the habits that actually turn knowledge into marks on the day.

Strategy 1 — When stuck, test the four options

The answer is always one of the four. If the algebra looks scary, try each option in the original question and see which one fits. Start with the simplest-looking one.

Example: Solve log2(x+2)+log2(x4)=4\log_2(x+2) + \log_2(x-4) = 4. Options: 4, 2, 6, 12-4,\ 2,\ 6,\ 12.

Strategy 2 — Put in easy numbers for the letters

If a question is all abstract letters, replace them with small numbers and see what happens.

Example: "Which expression is always even when nn is odd?" — just set n=1n = 1 and test each option. If two still survive, try n=3n = 3.

Strategy 3 — The 3-Pass System

Strategy 4 — Read the options first

The options themselves are clues:

Strategy 5 — The classic traps (know them, dodge them)

The tempting-but-wrong move Why it is wrong Do this instead
a2+b2=a+b\sqrt{a^2 + b^2} = a + b You cannot split a root across a + Leave it as a2+b2\sqrt{a^2+b^2} unless it simplifies exactly
Down 10% then up 10% = back to start It is actually a 1% loss Multiply the multipliers: 0.9×1.1=0.990.9 \times 1.1 = 0.99
log(a+b)=loga+logb\log(a + b) = \log a + \log b Only products split log(ab)=loga+logb\log(ab) = \log a + \log b
x+ax+b=ab\dfrac{x + a}{x + b} = \dfrac{a}{b} You cannot cancel across a + Only cancel things that are multiplied
(x)2=x2(-x)^2 = -x^2 Squaring always gives a positive (x)2=+x2(-x)^2 = +x^2

Strategy 6 — Smart guessing

If you genuinely have no idea:


Part 4 — Your Study Plan

You do not need to study everything. You need to study the right things, consistently. Little and often beats one long panic. This plan is built around this book — every step names exactly what to do with it.

Six to four weeks before

Three weeks before

Two weeks before

One week before

The night before


Priority List — If You Are Short on Time

If you cannot cover everything, study in this order — the topics near the top appear most often and are the most learnable:

  1. Fractions, decimals, and percentages → Chapter 2
  2. Basic algebra (factorising, simplifying, solving equations) → Chapter 1
  3. Functions and graphs (especially parabolas and straight lines) → Chapter 3
  4. Exponents and logarithms → Chapter 1 (Sections 1.5–1.6)
  5. Trigonometry (special angles, SOHCAHTOA, CAST) → Chapter 4
  6. Analytical geometry (distance, midpoint, gradient) → Chapter 5 (Section 5.8)
  7. Probability and counting → Chapter 6
  8. Sequences and series → Chapter 8
  9. Financial maths → Chapter 7
  10. Geometry and measurement → Chapter 5
  11. Data and statistics → Chapter 6

However short your time is, protect one full mock exam (Chapter 9) — pacing a real 3-hour paper is a skill of its own, and the first time you practise it must not be on test day.


A Final Word

When you sit down to write the NBT MAT, you will hit questions that look unlike anything you have practised. That is done on purpose — the examiners want to see how you think, not just what you have memorised.

So when a question looks strange, do not panic. Take a slow breath. Read it again, in plain words. Look at the four options for clues. Break the problem into smaller pieces. Ask yourself: what simple idea is hiding underneath all this wording? Almost always, the scary-looking question is just a familiar idea wearing a costume.

You have prepared. You understand the ideas, not just the tricks. Trust that.

Work calmly. Show your steps. Answer every single question — even the ones you guess.

You can do this.

From The Complete NBT Mathematics Practice Book by Eben Roux — download the full 796-page PDF. Free to share, complete and unaltered.