Exponential and logarithmic models AI HL

Most questions here put an unknown in an exponent or ask for the constants of a model, and the calculator does the arithmetic once the equation is set up. Rational exponents turn roots into powers, logarithms bring an unknown power down to ground level, and taking logarithms of data turns a power or exponential relationship into a straight line that regression can fit. Work through the four sections in order, because each one uses the one before.

Formulas at a glance

The booklet lists the link between exponents and logarithms and the three laws of logarithms. The index laws, rational exponents, the model forms, the half-life constant and the linearised forms are not in the booklet, so learn them.

Exponent and logarithm (the base is 10 or e)

ax=b⟺x=loga⁡ba^x=b\iff x=\log_a b

In the booklet

Product law

loga⁡(xy)=loga⁡x+loga⁡y\log_a(xy)=\log_a x+\log_a y

In the booklet

Quotient law

loga⁡xy=loga⁡x−loga⁡y\log_a\dfrac{x}{y}=\log_a x-\log_a y

In the booklet

Power law

loga⁡xm=mloga⁡x\log_a x^m=m\log_a x

In the booklet

Rational exponents

am/n=(an)ma^{m/n}=\left(\sqrt[n]{a}\right)^m, a−m/n=1am/na^{-m/n}=\dfrac{1}{a^{m/n}}

Not in the booklet

Index laws

am×an=am+na^m\times a^n=a^{m+n}, aman=am−n\dfrac{a^m}{a^n}=a^{m-n}, (am)n=amn(a^m)^n=a^{mn}

Not in the booklet

Exponential model

f(x)=kax+cf(x)=ka^x+c or f(x)=kerx+cf(x)=ke^{rx}+c

Not in the booklet

Natural logarithmic model

f(x)=a+bln⁡xf(x)=a+b\ln x

Not in the booklet

Decay constant from the half-life

k=ln⁡2half-lifek=\dfrac{\ln2}{\text{half-life}}

Not in the booklet

Power model, linearised

y=axn⇒log⁡y=nlog⁡x+log⁡ay=ax^n\Rightarrow\log y=n\log x+\log a

Not in the booklet

Exponential model, linearised

y=abx⇒log⁡y=(log⁡b)x+log⁡ay=ab^x\Rightarrow\log y=(\log b)x+\log a

Not in the booklet

Which method to use

Look at where the unknown is: in an exponent, inside a logarithm, or in the constants of a model. Then choose between a calculator graph and taking logarithms by hand.

The question hasUseNotes
A fractional power such as am/na^{m/n} to evaluate or simplifyThe root first, then the power: am/n=(an)ma^{m/n}=\left(\sqrt[n]{a}\right)^mA negative exponent means the reciprocal, not a negative answer
Powers of the same base multiplied, divided or raised to a powerAdd, subtract or multiply the exponentsWrite each root as a power first so that the laws apply
An unknown power with the exponential term alone, such as ax=ba^x=b or kerx=cke^{rx}=cTake ln\ln (or log10\log_{10}) of both sides and bring the power downIsolate the exponential term before taking logarithms
An unknown both in an exponent and outside it, such as 3x=10−x3^x=10-xGraph both sides and use the intersection toolLogarithms cannot separate the two places xx appears, and there may be more than one crossing
A logarithm of a product, quotient or powerThe three laws of logarithms, in base 10 or eeThe number inside a logarithm must be positive
A growth or decay model with a half-life, a constant to find or a time to reach a valueSet the model equal to the target, isolate the exponential term and take ln\lnFor a half-life the amount has fallen to half of the starting amount
A model a+bln⁡xa+b\ln x with two known valuesSubstitute each value to make two equations and solve them togetherThe model is only defined for x>0x>0
Data that curve, and the question asks whether the relationship is a power or an exponentialTake logarithms and fit a line: log⁡y\log y against log⁡x\log x for a power, log⁡y\log y against xx for an exponentialThe straight line with rr closest to 1 or −1-1 shows which relationship fits

Section 1 of 4 AHL1.10

Rational exponents

A rational exponent combines a root and a power: am/n=(an)ma^{m/n}=\left(\sqrt[n]{a}\right)^{m}, so 272/3=927^{2/3}=9 because the cube root of 2727 is 33 and 32=93^{2}=9. Negative exponents give reciprocals, and the usual index laws still apply to fractional powers. It covers simplifying numerical and algebraic expressions.

The index laws you need

These aren't a separate formula-booklet entry - they follow directly from the integer index laws you already know, just extended to fractional and negative powers.

Fractional powers

a1/n=ana^{1/n} = \sqrt[n]{a},   am/n=(an)m=amna^{m/n} = (\sqrt[n]{a})^m = \sqrt[n]{a^m}

The denominator nn is the root, the numerator mm is the power. Take the root first to keep the numbers small before you raise to the power.

Negative powers

a−m/n=1am/na^{-m/n} = \dfrac{1}{a^{m/n}}

A negative exponent means "reciprocal", not "negative answer". Combine with the fractional-power rule above once the sign is dealt with.

Key terms

What is the product law for indices?

When multiplying powers of the same base, add the exponents: am×an=am+na^{m}\times a^{n}=a^{m+n}. This holds for fractional exponents too, so use a common denominator.

e.g. x1/2×x1/3=x3/6+2/6=x5/6x^{1/2}\times x^{1/3}=x^{3/6+2/6}=x^{5/6}.

What is the quotient law for indices?

When dividing powers of the same base, subtract the exponents: aman=am−n\dfrac{a^{m}}{a^{n}}=a^{m-n}.

e.g. x5/2x1/2=x2\dfrac{x^{5/2}}{x^{1/2}}=x^{2}.

What is the power of a power law?

To raise a power to another power, multiply the exponents: (am)n=amn\left(a^{m}\right)^{n}=a^{mn}.

e.g. (x4)1/2=x2\left(x^{4}\right)^{1/2}=x^{2} and (x2/3)3=x2\left(x^{2/3}\right)^{3}=x^{2}.

How do you evaluate a rational power without a calculator?

Take the root first, then apply the power. This keeps the numbers small and easy to handle.

e.g. 813/481^{3/4}: the fourth root of 8181 is 33, and 33=273^{3}=27.

Worked example, step by step

The question Calculator needed [5 marks]

The time TT days that a planet takes to orbit the Sun is modelled by T=0.2d3/2,T=0.2d^{3/2}, where dd is the distance of the planet from the Sun, in millions of kilometres.

(a) Find the orbital time of a planet that is 58 million km from the Sun.

(b) Another planet takes 365 days to orbit the Sun. Find its distance from the Sun.

Worked solution: worked example

Answer: (a) 88.388.3 days (b) 149149 million km

Step 1 - Substitute the distance.

For (a), put d=58d=58 into the model: T=0.2×583/2.T=0.2\times58^{3/2}. M1 - Substitute d=58d=58 into T=0.2d3/2T=0.2d^{3/2}

Why? The exponent 32\tfrac32 means the square root cubed, (58)3\left(\sqrt{58}\right)^3. The calculator takes the power directly, as long as the whole exponent is in brackets.

Step 2 - Evaluate with the calculator.

Enter 0.2×58(3/2)0.2\times58^{(3/2)}. The calculator shows 583/2=441.7…58^{3/2}=441.7\ldots, so T=88.3 days.T=88.3\text{ days}. A1 - 88.388.3 days

Why? Without the brackets the calculator reads 58358^3 and then divides by 2, which gives a very different answer.

Step 3 - Isolate the power.

For (b), put T=365T=365: 365=0.2d3/2⇒d3/2=3650.2=1825.365=0.2d^{3/2}\quad\Rightarrow\quad d^{3/2}=\dfrac{365}{0.2}=1825. M1 - Isolate the power: d3/2=1825d^{3/2}=1825

Why? The power of dd must stand alone before it can be undone, so divide by the 0.20.2 first.

Step 4 - Raise both sides to the reciprocal power.

Raise both sides to the power 23\tfrac23: (d3/2)2/3=d1=d⇒d=18252/3.\left(d^{3/2}\right)^{2/3}=d^{\,1}=d\quad\Rightarrow\quad d=1825^{2/3}. M1 - Raise both sides to the power 23\tfrac23

Why? Multiplying the exponents gives 32×23=1\tfrac32\times\tfrac23=1, so the power of dd is removed. The reciprocal of 32\tfrac32 is 23\tfrac23.

Step 5 - Evaluate with the calculator.

Enter 1825(2/3)1825^{(2/3)}. The calculator shows d=149.3…d=149.3\ldots, so the distance is d=149 million km.d=149\text{ million km}. A1 - 149149 million km

Why? Check with the unrounded value: 0.2×149.33/2=3650.2\times149.3^{3/2}=365, which agrees with the given time.

Marks (5):
M1 Substitute d=58d=58 into T=0.2d3/2T=0.2d^{3/2}
A1 88.388.3 days
M1 Isolate the power: d3/2=1825d^{3/2}=1825
M1 Raise both sides to the power 23\tfrac23
A1 149149 million km

Now you try: enter your answers to be marked

These are real practice questions. Type your answer in each box and press Mark it. The worked solution and mark scheme appear straight away, and your first attempt is the one that counts. Use the keyboard for powers, roots and trig, and the working box for your notes.

1
Easy
Calculator
[3 marks]
Kleiber's law models an animal's metabolic rate RR (in watts) in terms of its mass mm (in kg) as R=4m3/4.R = 4m^{3/4}. Find the metabolic rate of an animal with mass 1616 kg.
[3]
Fractions & powers
Other

Worked solution: practice question 1

2
Medium
Calculator
[3 marks]
The radius rr of a sphere with volume VV is given by r=(3V4π)1/3.r = \left(\dfrac{3V}{4\pi}\right)^{1/3}. Find rr when V=288πV = 288\pi cm3^{3}, giving your answer as an exact value.
[3]
Fractions & powers
Other

Worked solution: practice question 2

3
Easy
No calc
[4 marks]
Simplify each expression without a calculator.
(a)  82/38^{2/3}
[1]
(b)  16−3/416^{-3/4}
[2]
(c)  x1/2⋅x3/2x^{1/2} \cdot x^{3/2}
[1]
Fractions & powers
Other

Worked solution: practice question 3

Your score so far: 0 out of 10 marks

Common mistakes

  • Treating a negative exponent as a negative value. 81−1/281^{-1/2} is not −811/2-81^{1/2} - the minus sign in the exponent means "take the reciprocal", so the answer for a positive base is always positive.
  • Adding the exponents wrongly. Multiplying x1/4x^{1/4} by x2/3x^{2/3} means adding 14+23=1112\tfrac14+\tfrac23=\tfrac{11}{12}, not adding the tops and the bottoms to get 37\tfrac37. Use a common denominator before adding.
  • Rooting last instead of first. Computing am/na^{m/n} as amn\sqrt[n]{a^m} means raising to a large power before rooting, which produces huge numbers to simplify by hand; (an)m(\sqrt[n]{a})^m keeps the arithmetic manageable.

Try some questions on rational exponents

Section 2 of 4 AHL1.9 AHL2.9

Solving exponential/log equations

To solve an equation with the unknown in an exponent, first isolate the exponential term, then take logarithms of both sides and use ln⁡(ax)=xln⁡a\ln\left(a^{x}\right)=x\ln a to bring the exponent down. When the unknown is also outside the exponent, as in 3x=10−x3^{x}=10-x, the calculator graphs both sides and finds the crossings. It also covers checking solutions in context.

Taking logs to solve

There are two cases: the exponential term can be made to stand alone, or the unknown also appears outside the exponent.

Isolate, then take logs

ax=b⇒x=ln⁡bln⁡aa^x=b \Rightarrow x=\dfrac{\ln b}{\ln a}

Get the exponential term alone on one side first, then apply ln\ln (or log\log) to both sides and use the power law to bring the exponent down.

Unknown inside and outside the exponent

3x=10−x3^x=10-x

Logarithms cannot separate the two places xx appears. Graph both sides, then use the intersection tool and read the xx-coordinate of each crossing.

Key terms

What is a logarithm?

The logarithm loga⁡b\log_{a}b is the power to which aa must be raised to give bb. It undoes an exponential with the same base.

e.g. log10⁡1000=3\log_{10}1000=3 because 103=100010^{3}=1000.

What is the natural logarithm?

The natural logarithm ln⁡x\ln x is the logarithm to base ee, and it is the inverse of exe^{x}. It is only defined for x>0x>0.

e.g. ln⁡(e5)=5\ln\left(e^{5}\right)=5 and eln⁡7=7e^{\ln7}=7.

What is the power law for logarithms?

Like the power law in the booklet, ln⁡(ax)=xln⁡a\ln\left(a^{x}\right)=x\ln a. It lets you bring an unknown exponent down so it can be solved for.

e.g. ln⁡(210)=10ln⁡2\ln\left(2^{10}\right)=10\ln2.

Why must you isolate the exponential term first?

The logarithm must be applied to the whole of one side, so any added or multiplied numbers need to be removed before taking logs. Taking logs of a sum does not split into separate logs.

e.g. 3ex+2=143e^{x}+2=14 gives ex=4e^{x}=4, so x=ln⁡4≈1.39x=\ln4\approx1.39.

Worked example, step by step

The question Calculator needed [5 marks]

The noise level LL of a sound, in decibels (dB), is given by L=10log10⁡I,L=10\log_{10}I, where II is the intensity of the sound, measured as a multiple of the quietest sound that can be heard. One machine in a workshop has a noise level of 62 dB. When nn identical machines run together, the intensity of the sound is nn times the intensity of one machine.

(a) Show that the noise level of nn machines is 62+10log10⁡n62+10\log_{10}n dB.

(b) Find the smallest number of machines that gives a noise level of at least 80 dB.

Worked solution: worked example

Answer: (a) shown (b) 6464 machines

Step 1 - Write the level for nn machines.

Let II be the intensity of one machine, so 10log10⁡I=6210\log_{10}I=62. The intensity of nn machines is nInI, so L=10log10⁡(nI).L=10\log_{10}(nI). M1 - Write L=10log10⁡(nI)L=10\log_{10}(nI) and apply the product law

Why? The question says the intensity is multiplied by nn, so the number inside the logarithm is nInI.

Step 2 - Apply the product law.

L=10(log10⁡n+log10⁡I)=10log10⁡n+10log10⁡I=10log10⁡n+62.L=10\left(\log_{10}n+\log_{10}I\right)=10\log_{10}n+10\log_{10}I=10\log_{10}n+62. So the noise level is 62+10log10⁡n62+10\log_{10}n dB. A1 - Complete the working to 62+10log10⁡n62+10\log_{10}n

Why? The product law loga⁡xy=loga⁡x+loga⁡y\log_a xy=\log_a x+\log_a y separates nn from II, and the part with II is the known 6262 dB.

Step 3 - Set up the inequality for (b).

62+10log10⁡n≥80⇒log10⁡n≥1.8.62+10\log_{10}n\ge80\quad\Rightarrow\quad\log_{10}n\ge1.8. M1 - Form 62+10log10⁡n≥8062+10\log_{10}n\ge80, or log10⁡n≥1.8\log_{10}n\ge1.8

Why? Subtract 6262 and divide by 1010 to leave the logarithm alone. The level must be at least 8080, so the sign is ≥\ge.

Step 4 - Solve with the calculator.

Rewrite as a power of 10: n≥101.8n\ge10^{1.8}. Enter 101.810^{1.8} to get n≥63.1.n\ge63.1. A1 - n≥63.1n\ge63.1

Why? A logarithm to base 10 is undone by a power of 10. The calculator shows 63.09…63.09\ldots

Step 5 - Choose the whole number.

The number of machines is a whole number and must be at least 63.163.1, so n=64n=64. Check: 62+10log10⁡63=79.99<8062+10\log_{10}63=79.99<80 but 62+10log10⁡64=80.06≥8062+10\log_{10}64=80.06\ge80. R1 - n=64n=64, since 6363 machines give a level below 8080 dB

Why? Rounding 63.163.1 down to 6363 would give a level just under 8080 dB, so it must be rounded up.

Marks (5):
M1 Write L=10log10⁡(nI)L=10\log_{10}(nI) and apply the product law
A1 Complete the working to 62+10log10⁡n62+10\log_{10}n
M1 Form 62+10log10⁡n≥8062+10\log_{10}n\ge80, or log10⁡n≥1.8\log_{10}n\ge1.8
A1 n≥63.1n\ge63.1
R1 n=64n=64, since 6363 machines give a level below 8080 dB

Now you try: enter your answers to be marked

1
Easy
Calculator
[2 marks]
Use your graphic display calculator to solve 2x=3x2^x = 3x for the larger positive root.
[2]
Fractions & powers
Other

Worked solution: practice question 1

2
Medium
Calculator
[2 marks]
pH =−log10⁡[H+].= -\log_{10}[H^{+}]. Find the pH when [H+]=6.5×10−4[H^{+}] = 6.5 \times 10^{-4} mol/L.
[2]
Fractions & powers
Logs
Other

Worked solution: practice question 2

3
Hard
Calculator
[4 marks]
Coffee cools as T(t)=20+70e−0.1tT(t) = 20 + 70e^{-0.1t} (°C, tt min).

Find when it reaches 50 °C.
[4]
Fractions & powers
Logs
Other

Worked solution: practice question 3

Your score so far: 0 out of 8 marks

Common mistakes

  • Taking logs before isolating the exponential. ln\ln can only simplify a power once the exponential term stands alone on one side - taking logs of 2ex+3=112e^x+3=11 directly does nothing useful.
  • Splitting the logarithm of a sum. ln⁡(x+3)\ln(x+3) is not ln⁡x+ln⁡3\ln x+\ln3. The laws only split a product, a quotient or a power, so a sum has to be dealt with another way.
  • Stopping at the first crossing. When the calculator graphs both sides, look at the whole graph before reading an answer: ex=4xe^{x}=4x has two solutions, x=0.357x=0.357 and x=2.15x=2.15, and a question may want either or both.

Try some questions on solving exponential/log equations

Section 3 of 4 AHL2.9

Exponential and log models

Exponential models such as f(x)=kax+cf(x)=ka^{x}+c describe growth, or decay towards the horizontal asymptote y=cy=c, while natural logarithmic models f(x)=a+bln⁡xf(x)=a+b\ln x describe a quantity that rises quickly and then slows down. Half-life is the time taken for an amount to halve. It covers choosing a model, interpreting its constants and calculating half-life.

The two model forms

An exponential model is f(x)=kax+cf(x)=ka^x+c or f(x)=kerx+cf(x)=ke^{rx}+c, and a natural logarithmic model is f(x)=a+bln⁡xf(x)=a+b\ln x. Neither form is in the formula booklet, so learn them.

Key terms

What is half-life?

Half-life is the time it takes for a decaying quantity to fall to half of its value. It stays the same wherever you start on the decay curve.

e.g. A sample of 120120 g with a half-life of 33 hours has 6060 g after 33 hours and 3030 g after 66 hours.

What do kk and aa represent in kax+cka^{x}+c?

If a>1a>1 the model shows growth and if 0<a<10<a<1 it shows decay, while kk scales the curve. At x=0x=0 the value is f(0)=k+cf(0)=k+c.

e.g. T(t)=60×0.9t+20T(t)=60\times0.9^{t}+20 has a=0.9<1a=0.9<1, so it decays, and T(0)=80T(0)=80.

What is the horizontal asymptote of an exponential model?

The asymptote is the line y=cy=c, a value the graph gets closer to but never reaches. It often represents a limiting value such as room temperature.

e.g. The graph of f(x)=5×0.5x+3f(x)=5\times0.5^{x}+3 approaches y=3y=3 as xx becomes large.

What does the sign of r tell you in erxe^{rx}?

In a model kerxke^{rx}, a positive rr means growth and a negative rr means decay. The larger |r||r| is, the faster the change.

e.g. N=200e0.05tN=200e^{0.05t} grows, but N=200e−0.05tN=200e^{-0.05t} decays.

Worked example, step by step

The question Calculator needed [6 marks]

A culture contains 250 bacteria at the start of an experiment, and 410 bacteria after 3 hours. The number of bacteria NN after tt hours is modelled by N=250ert.N=250e^{rt}.

(a) Find the value of rr.

(b) Find the time for the number of bacteria to double.

(c) Find the number of bacteria after 10 hours.

Worked solution: worked example

Answer: (a) 0.1650.165 (b) 4.204.20 hours (c) 13001300

Step 1 - Use the data point.

When t=3t=3, N=410N=410: 410=250e3r⇒e3r=410250=1.64.410=250e^{3r}\quad\Rightarrow\quad e^{3r}=\dfrac{410}{250}=1.64. M1 - Substitute t=3t=3 and N=410N=410, and isolate the exponential term

Why? Substitute the known pair into the model and isolate the exponential term, so that a logarithm can undo it.

Step 2 - Solve for rr.

Take natural logarithms: 3r=ln⁡1.643r=\ln1.64, so r=ln⁡1.643=0.165.r=\dfrac{\ln1.64}{3}=0.165. Store the unrounded value 0.16489…0.16489\ldots for the next parts. A1 - r=0.165r=0.165

Why? ln\ln is the inverse of exe^{x}, so ln⁡(e3r)=3r\ln\left(e^{3r}\right)=3r. Keeping all the digits stops rounding errors building up.

Step 3 - Write the doubling condition.

The number doubles when N=500N=500: 500=250ert⇒ert=2.500=250e^{rt}\quad\Rightarrow\quad e^{rt}=2. M1 - Set N=500N=500 and isolate ert=2e^{rt}=2

Why? The starting amount 250250 cancels, so the doubling time does not depend on where the doubling starts. This is the growth version of a half-life.

Step 4 - Solve for tt.

rt=ln⁡2⇒t=ln⁡2r=0.6931…0.16489…=4.20 hours.rt=\ln2\quad\Rightarrow\quad t=\dfrac{\ln2}{r}=\dfrac{0.6931\ldots}{0.16489\ldots}=4.20\text{ hours}. A1 - 4.204.20 hours

Why? Take ln\ln of both sides, then divide by rr. Using the stored value of rr avoids a rounding slip.

Step 5 - Predict the number after 10 hours.

N=250e10r=250e1.6489…=1300 bacteria.N=250e^{10r}=250e^{1.6489\ldots}=1300\text{ bacteria}. M1 - Substitute t=10t=10 into N=250ertN=250e^{rt} A1 - 13001300 bacteria

Why? Substitute t=10t=10 with the stored rr. The calculator shows 1300.4…1300.4\ldots, and the answer is a whole number of bacteria to 3 significant figures.

Marks (6):
M1 Substitute t=3t=3 and N=410N=410, and isolate the exponential term
A1 r=0.165r=0.165
M1 Set N=500N=500 and isolate ert=2e^{rt}=2
A1 4.204.20 hours
M1 Substitute t=10t=10 into N=250ertN=250e^{rt}
A1 13001300 bacteria

Now you try: enter your answers to be marked

1
Easy
Calculator
[3 marks]
The improvement in a typist's score after hh hours of practice (h≥1h \ge 1) is modelled by S(h)=50+12ln⁡(h).S(h) = 50 + 12\ln(h).
(a) State the score after 1 hour of practice.
[1]
(b) Find the score after 10 hours of practice.
[2]
Fractions & powers
Logs
Other

Worked solution: practice question 1

2
Medium
Calculator
[6 marks]
A radioactive sample decays according to m(t)=80e−ktm(t) = 80e^{-kt} grams, where tt is measured in years. After 5 years, 45 grams of the sample remain.
(a) Find the value of kk.
[3]
(b) Find the half-life of the sample.
[3]
Fractions & powers
Logs
Other

Worked solution: practice question 2

3
Hard
Calculator
[2 marks]
On the Richter scale, magnitude differs by log10\log_{10} of amplitude ratio. How many times stronger (amplitude) is a magnitude 6.5 quake than a magnitude 5.0 one?
[2]
Fractions & powers
Logs
Other

Worked solution: practice question 3

Your score so far: 0 out of 11 marks

Common mistakes

  • Forgetting the domain of a log model. A natural log model f(x)=a+bln⁡xf(x)=a+b\ln x is undefined for x≤0x\le0 - always state x>0x>0 when asked for the domain.
  • Treating the asymptote cc as zero. In f(x)=kax+cf(x)=ka^x+c, the constant cc shifts the whole curve vertically - a decay model levels off at y=cy=c, not at y=0y=0, unless the question says otherwise.
  • Treating a logarithmic scale as a straight scale. On a scale such as the Richter scale a difference of 22 means the amplitude is multiplied by 102=10010^{2}=100, not by 22. Work out the power of 10 first.

Try some questions on exponential and log models

Section 4 of 4 AHL2.10

Linearising data with logarithms

Linearising data means taking logarithms so that a curved relationship becomes a straight line. For an exponential model y=abxy=ab^{x}, plot log⁡y\log y against xx; for a power model y=axny=ax^{n}, plot log⁡y\log y against log⁡x\log x. The gradient and intercept of the line then give the parameters of the original model.

Two linearising transforms

In examinations, students are not expected to sketch log-log or semi-log graphs by hand.

Power model y=axny=ax^n

log⁡y=nlog⁡x+log⁡a\log y = n\log x + \log a

Plot log⁡y\log y against log⁡x\log x (a log-log graph). The line's gradient gives nn; its yy-intercept gives log⁡a\log a, so a=10intercepta=10^{\text{intercept}}.

Exponential model y=abxy=ab^x

log⁡y=(log⁡b)x+log⁡a\log y = (\log b)x + \log a

Plot log⁡y\log y against xx itself (a semi-log graph). The gradient gives log⁡b\log b; the yy-intercept gives log⁡a\log a.

Key terms

What does linearising data mean?

It means transforming the variables, usually with logarithms, so that the points lie close to a straight line. A straight line can then be fitted and its equation converted back to the original model.

e.g. Taking logs of y=3×2xy=3\times2^{x} gives log⁡y=log⁡3+xlog⁡2\log y=\log3+x\log2, which is linear in xx.

What are the gradient and intercept for y=abxy=ab^{x}?

Plotting log⁡y\log y against xx gives a line with gradient log⁡b\log b and vertical intercept log⁡a\log a. The parameters are recovered by undoing the logarithm.

e.g. For y=5×10xy=5\times10^{x}, log⁡y=log⁡5+x\log y=\log5+x, so the gradient is 11 and the intercept is log⁡5\log5.

What is a semi-log graph?

A semi-log graph has a logarithmic scale on one axis only, or equivalently plots log⁡y\log y against xx. A straight line on it indicates an exponential relationship.

e.g. If log⁡y\log y against xx is a straight line, then y=abxy=ab^{x} for some aa and bb.

What is a log-log graph?

A log-log graph plots log⁡y\log y against log⁡x\log x. A straight line on it indicates a power relationship, and the gradient is the power.

e.g. If the line has gradient 22, then y=ax2y=ax^{2} for some constant aa.

Worked example, step by step

The question Calculator needed [7 marks]

The number of downloads of a new app in week ww after its release is DD thousand. The recorded values (w,D)(w, D) are (1,90)(1, 90), (2,135)(2, 135), (3,203)(3, 203), (4,304)(4, 304) and (5,456)(5, 456). The downloads are modelled by D=abwD=ab^w.

(a) Show that ln⁡D\ln D is a linear function of ww, and state its gradient and its intercept in terms of aa and bb.

(b) Use linear regression of ln⁡D\ln D on ww to find the values of aa and bb.

(c) Predict the number of downloads, in thousands, in week 7.

Worked solution: worked example

Answer: (a) gradient ln⁡b\ln b, intercept ln⁡a\ln a (b) a=60.0a=60.0, b=1.50b=1.50 (c) 10301030 thousand

Step 1 - Take logarithms of the model.

ln⁡D=ln⁡(abw)=ln⁡a+ln⁡(bw)=ln⁡a+wln⁡b.\ln D=\ln\left(ab^{w}\right)=\ln a+\ln\left(b^{w}\right)=\ln a+w\ln b. M1 - Take ln\ln of both sides and use the product and power laws

Why? The product law splits abwab^w into ln⁡a+ln⁡bw\ln a+\ln b^w, and the power law brings the exponent ww down in front.

Step 2 - State the straight line.

ln⁡D=(ln⁡b)w+ln⁡a,\ln D=(\ln b)\,w+\ln a, a straight line when ln⁡D\ln D is plotted against ww, with gradient ln⁡b\ln b and intercept ln⁡a\ln a. A1 - ln⁡D=(ln⁡b)w+ln⁡a\ln D=(\ln b)w+\ln a, with gradient ln⁡b\ln b and intercept ln⁡a\ln a

Why? It has the form y=mx+cy=mx+c with y=ln⁡Dy=\ln D and x=wx=w. A straight line against ww is the sign of an exponential relationship.

Step 3 - Fit the line with the calculator.

Enter ww in one list, and in a second list enter ln⁡D\ln D, which is 4.50,4.91,5.31,5.72,6.124.50, 4.91, 5.31, 5.72, 6.12. Linear regression gives ln⁡D=0.406w+4.09.\ln D=0.406w+4.09.

Why? The calculator can fill the second list by applying ln\ln to the list of DD values. Keep the unrounded gradient and intercept for the next step.

Step 4 - Undo the logarithms.

The gradient is ln⁡b\ln b and the intercept is ln⁡a\ln a, so b=e0.4057…=1.50,a=e4.0944…=60.0.b=e^{0.4057\ldots}=1.50,\qquad a=e^{4.0944\ldots}=60.0. M1 - Use b=egradientb=e^{\text{gradient}} and a=eintercepta=e^{\text{intercept}} A1 - b=1.50b=1.50 A1 - a=60.0a=60.0

Why? The regression gives logarithms of the constants, so raise ee to each value to recover bb and aa.

Step 5 - Predict the downloads in week 7.

D=60.0×1.507≈1030 thousand.D=60.0\times1.50^{7}\approx1030\text{ thousand}. M1 - Substitute w=7w=7 into D=abwD=ab^w with their aa and bb A1 - 10301030 thousand

Why? Use the unrounded values of aa and bb, which give 1027…1027\ldots. Week 7 is outside the data (weeks 1 to 5), so this is an extrapolation and less reliable.

Marks (7):
M1 Take ln\ln of both sides and use the product and power laws
A1 ln⁡D=(ln⁡b)w+ln⁡a\ln D=(\ln b)w+\ln a, with gradient ln⁡b\ln b and intercept ln⁡a\ln a
M1 Use b=egradientb=e^{\text{gradient}} and a=eintercepta=e^{\text{intercept}}
A1 b=1.50b=1.50
A1 a=60.0a=60.0
M1 Substitute w=7w=7 into D=abwD=ab^w with their aa and bb
A1 10301030 thousand

Now you try: enter your answers to be marked

1
Easy
No calc
[3 marks]
A student writes the linearised form of y=axny = ax^n as ln⁡y=nln⁡x+ln⁡a.\ln y = n \ln x + \ln a.
(a) State whether this is valid.
[1]
(b) State what the gradient represents.
[1]
(c) State what the yy-intercept represents.
[1]
Fractions & powers
Logs
Other

Worked solution: practice question 1

2
Medium
Calculator
[3 marks]
After linearising y=axny = ax^n, a line of best fit is found: log10⁡y=2.1log10⁡x+0.6.\log_{10} y = 2.1 \log_{10} x + 0.6. Find aa and n.n.
(a)(i) Find nn.
[1]
(a)(ii) Find aa.
[2]
Fractions & powers
Logs
Other

Worked solution: practice question 2

3
Hard
Calculator
[6 marks]
Data: (x,y)(x, y) =
xx13927
yy1.57.840.5210
. Propose a suitable model, linearise the data, find aa and nn by regression, and predict yy at x=6.x = 6.
(a)(i) Find aa.
[2]
(a)(ii) Find nn.
[2]
(a)(iii) Predict yy at x=6x=6.
[2]
Fractions & powers
Logs
Other

Worked solution: practice question 3

Your score so far: 0 out of 12 marks

Common mistakes

  • Using the wrong pair of variables. A power model needs log⁡y\log y plotted against log⁡x\log x; an exponential model needs log⁡y\log y plotted against plain xx. Mixing these up gives a curve that still isn't straight.
  • Reading the gradient as aa and the intercept as nn (or vice versa). For y=axny=ax^n, the gradient of the linearised line is the power nn; the constant aa only appears after undoing the log on the intercept.
  • Forgetting to undo the log on the intercept. The yy-intercept of the linearised line is log⁡a\log a, not aa itself, so raise 10 (or ee, depending on which logarithm was used) to that value to recover aa. For y=abxy=ab^x the gradient is log⁡b\log b, so bb is recovered the same way.

Try some questions on linearising data with logarithms

Using your GDC

The exact key presses for the TI-84 Plus CE, Casio fx-CG50, Casio fx-CG100 and TI-Nspire CX, with what the screen shows. Pick your calculator.

Use brackets, powers and roots correctly

Used in: Rational exponents

How to use brackets, powers and roots correctly on a TI-84 Plus CE

A fraction: brackets round the top and the bottom

The calculator does powers and roots first, then multiplication and division, then addition and subtraction, and anything in brackets before all of these. For 14.7+6.32.5−1.4\dfrac{14.7+6.3}{2.5-1.4}, put brackets round the top and round the bottom: ( 1 4 . 7 + 6 . 3 ) ÷ ( 2 . 5 − 1 . 4 ) enter.

19.119.1 to 3 significant figures.

(14.7+6.3)/(2.5-1.4)
19.09090909

The fraction template does the bracketing for you. Press math, then ▶ to the FRAC list, and choose 1:n/d. Type the top, press ▼, type the bottom, then press ▶ to leave the fraction before typing anything else.

Powers: leave the exponent, and bracket a negative base

Type the base, press ^, type the exponent, then press ▶ to come back down to the line before typing anything else. For 1000(1+0.0454)121000\left(1+\dfrac{0.045}{4}\right)^{12} press 1 0 0 0 ( 1 + 0 . 0 4 5 ÷ 4 ) ^ 1 2 enter. The brackets round 1+0.04541+\dfrac{0.045}{4} make the whole of it the base.

1143.671143.67 to 2 decimal places.

1000(1+0.045/4)^12
1143.674441

When the exponent has more than one number in it, bracket it. For 272/327^{2/3} press 2 7 ^ ( 2 ÷ 3 ) enter. Without the brackets, 27^2/3 means (272)÷3=243(27^2)\div 3 = 243.

272/3=927^{2/3} = 9.

27^(2/3)
9

Squaring is done before the negative sign, so a negative base needs brackets. ( (−) 3 ) x² enter gives 99, but (−) 3 x² enter gives −9-9.

(-3)²
9
-3²
-9
Roots: everything under the root goes inside it

Press 2nd x² (√) and type everything that is under the root: for 6.42+3.62\sqrt{6.4^2+3.6^2} press 6 . 4 x² + 3 . 6 x², then ▶ to close the root, then enter. In Classic mode press ) to close the root instead.

7.347.34 to 3 significant figures.

√(6.4²+3.6²)
7.3430239

For a cube root press math and choose 4:∛(. For any other root write it as a power with a bracketed exponent: the fifth root of xx is x^(1/5).

When in doubt, add brackets. An extra pair never changes an answer, but a missing pair does.

The full guide: Use brackets, powers and roots correctly

Solve an equation numerically (including multiple solutions)

Used in: Solving exponential/log equations

How to solve an equation numerically on a TI-84 Plus CE

Numeric Solver (one solution per starting value)

Press math, then ▲ to jump to the bottom of the menu, and choose C:Numeric Solver.

Enter the equation as E1 = E2. Type X,T,θ,n ^ 3 − 3 X,T,θ,n for E1 and 1 for E2, then choose OK. The equation does not need rearranging to equal zero.

E1=X^3-3X
E2=1

Put the cursor on X and type a number close to the solution you want, for example 2. Leave the bounds as they are ({-1E99,1E99}) unless you want to search only a smaller interval.

X=2
bound={-1E99,1E99}

Press alpha enter (SOLVE). The solution is marked with a small square, and E1-E2 shows how far the two sides are apart: it should be 0 or a very small number.

x=1.879x = 1.879 to 3 decimal places, which is 1.881.88 to 3 significant figures.

X=1.879385
E1-E2=0

The solver finds the solution nearest your starting value. For the others, start again with another value: (−) 1 finds x=−1.53x = -1.53 and 0 finds x=−0.347x = -0.347.

X=-1.532089
X=-0.347296
Another way: from the graph

Graph the left side as Y1 and the right side as Y2, then use 2nd trace 5:intersect at each crossing. The graph also shows how many solutions there are, which tells you how many starting values to try. The steps are in find where two graphs meet.

A starting value near the solution you want is what matters. If the solver gives a different solution from the one you expected, look at the graph and start closer.

The full guide: Solve an equation numerically (including multiple solutions)

Find where two graphs meet (intersection)

Used in: Solving exponential/log equations

How to find where two graphs meet on a TI-84 Plus CE

Graph both and use intersect

Press Y=, clear any old functions with clear, then type X,T,θ,n x² − 3 next to Y1=. Press ▼ and type 2 X,T,θ,n + 1 next to Y2=.

Plot1 Plot2 Plot3
\Y1=X²-3
\Y2=2X+1
\Y3=

Press window and set Xmin = −3, Xmax = 5, Ymin = −4, Ymax = 10. For a negative number use the (−) key, not the subtraction key. Press graph. Both curves should show, with two crossing points on the screen.

WINDOW
Xmin=-3
Xmax=5
Xscl=1
Ymin=-4
Ymax=10
Yscl=1

Press 2nd trace (CALC) and choose 5:intersect. First curve? appears with the cursor on one graph: press enter. Second curve? appears: press enter again. If more than two graphs are drawn, use ▲ or ▼ to put the cursor on the right graph before pressing enter.

Second curve?

Guess? appears. Use ◀ or ▶ to move the cursor close to the right-hand crossing, then press enter.

The screen shows Intersection with x=3.24x = 3.24 and y=7.47y = 7.47.

Intersection X=3.236068 Y=7.472136

For the other crossing, repeat 2nd trace 5:intersect, enter, enter, and this time move the guess close to the left-hand crossing before pressing enter.

x=−1.24x = -1.24 and y=−1.47y = -1.47.

Intersection X=-1.236068 Y=-1.472136

The guess decides which crossing the calculator finds, so give a guess near the point you want. A crossing outside the window cannot be found, so check WINDOW first.

The full guide: Find where two graphs meet (intersection)

Enter a calculation on the home screen

Used in: Exponential and log models

How to enter a calculation on a TI-84 Plus CE

Type a calculation and press enter

Go to the home screen: press 2nd mode (QUIT) from any menu or app. Press clear if the entry line is not empty.

Type the calculation as you would write it, then press enter. For 3(2.4)2−5×1.83(2.4)^2 - 5 \times 1.8 press 3 ( 2 . 4 ) x² − 5 × 1 . 8 enter.

The calculator understands a number next to a bracket as a multiplication, so × is not needed in 3(2.4)3(2.4). If a fraction, square root or power template is on the screen, press ▶ to move out of it before you carry on typing.

8.288.28

3(2.4)²-5*1.8
8.28
Use the last answer in the next calculation

To increase the answer by 15%, press × 1 . 1 5 enter on a blank line. The calculator puts Ans, the last answer, in front of the operator for you.

You can also type Ans yourself, anywhere in a calculation: press 2nd (−) (ANS), for example 2nd (−) (ANS) × 1 . 1 5 enter.

9.5229.522, which is 9.529.52 to 3 significant figures.

Ans*1.15
9.522

To use an earlier entry or answer, press ▲ to move up through the history, then enter to paste the highlighted one on the entry line. Press 2nd enter (ENTRY) to bring back your last entry and change it.

Ans always holds the last answer that was worked out, so if you do a different calculation in between, Ans changes. Type a number you need later, or keep it on the screen, rather than relying on Ans for many steps.

The full guide: Enter a calculation on the home screen

Linear regression (line of best fit) and r

Used in: Linearising data with logarithms

How to find a line of best fit and r on a TI-84 Plus CE

Turn on r (once)

Press mode, use ▼ to reach STAT DIAGNOSTICS, move to ON with ▶ and press enter. Then press 2nd mode (QUIT). With diagnostics off, the correlation coefficient rr is not shown. This setting stays on until you reset the calculator.

Enter the data and run LinReg(ax+b)

Press stat and choose 1:Edit. Clear any old values: move up to the L1 heading, press clear then enter, and do the same for L2.

Type the xx-values into L1, pressing enter after each: 3, 5, 6, 8, 9, 11, 12, 14. Press ▶ to move to L2 and type the yy-values, pressing enter after each: 41, 48, 47, 58, 60, 69, 66, 77. Both lists must have the same number of values.

L1 = 3, 5, 6, 8, 9, 11, 12, 14
L2 = 41, 48, 47, 58, 60, 69, 66, 77

Press stat, move to CALC with ▶, and choose 4:LinReg(ax+b). With the Stat Wizards on, a screen asks for Xlist and Ylist: set Xlist to L1 and Ylist to L2 (press 2nd 1 for L1 and 2nd 2 for L2), then move to Calculate and press enter. If the Stat Wizards are off, type LinReg(ax+b) L1,L2 and press enter.

LinReg(ax+b)
y=ax+b
a=3.2346939
b=30.7551020
r²=0.9678131
r=0.9837749

Read a (the gradient), b (the yy-intercept) and r.

y=3.23x+30.8y = 3.23x + 30.8 and r=0.984r = 0.984, both to 3 significant figures.

Predict a value from the line

Before you run LinReg, move to Store RegEQ on the wizard screen, press vars, then ▶ to Y-VARS and choose 1:Function then 1:Y1. The full-accuracy equation is then saved in Y1. On the home screen type Y1(10) and press enter.

y=63.1y = 63.1 when x=10x = 10, to 3 significant figures.

Y1(10)
63.1020408

If r does not appear, diagnostics are off: turn STAT DIAGNOSTICS on in the mode screen and run LinReg again. Use the full values of a and b, not rounded ones, when you predict.

The full guide: Linear regression (line of best fit) and r

Build and manipulate lists

Used in: Linearising data with logarithms

How to build and manipulate lists on a TI-84 Plus CE

Enter a list

Press stat and choose 1:Edit. If L1 already has numbers in it, press ▲ to put the cursor on the name L1, then press clear and enter to empty it.

Type the values into L1, one at a time, pressing enter after each: 12, 7, 15, 9, 11.

L1
12
7
15
9
11

You can also type the whole list on the home screen. Press 2nd ( ({), type the values separated by ,, then 2nd ) (}) sto▶ 2nd 1 (L1) and enter.

{12,7,15,9,11}→L1
{12 7 15 9 11}
Sort a list

On the home screen press 2nd stat (LIST), then ▶ to the OPS tab, and choose 1:SortA(. Type 2nd 1 (L1) ) and press enter.

The list in L1 is now in order, smallest first. 2:SortD( sorts it largest first.

The sorted list is 7, 9, 11, 12, 15.

SortA(L1)
Done
Apply an expression to the whole list

On the home screen type the expression with the list name in it, then store the result in another list. To square every value in L1 and store the squares in L2, press 2nd 1 (L1) x² sto▶ 2nd 2 (L2) and enter. The expression is applied to every value in the list at once.

The squares are 49, 81, 121, 144, 225.

L1²→L2
{49 81 121 144 225}
Add up a list

Press 2nd stat (LIST), then ▶ to the MATH tab and choose 5:sum(. Type 2nd 2 (L2) ) and press enter.

The sum of the squares is 620620. Using sum( on L1 gives the total of the values, 5454.

sum(L2)
620
Clear a list

Press stat and choose 4:ClrList. Type 2nd 1 (L1) and press enter. The list is now empty. To clear both lists, type 2nd 1 , 2nd 2 before pressing enter.

ClrList L1
Done

SortA( puts the list itself in order, so the original order is gone. If you need it again, store a copy in another list before you sort.

The full guide: Build and manipulate lists

See the full list of GDC guides for every calculator task.

Syllabus content covered

CodeSyllabus content
AHL1.9Laws of logarithms: loga⁡xy=loga⁡x+loga⁡y\log_a xy=\log_a x+\log_a y, loga⁡xy=loga⁡x−loga⁡y\log_a\dfrac{x}{y}=\log_a x-\log_a y, loga⁡xm=mloga⁡x\log_a x^m=m\log_a x for a,x,y>0a, x, y>0. In examinations, aa will equal 10 or ee.
AHL1.10Simplifying expressions, both numerically and algebraically, involving rational exponents.
AHL2.9Exponential models to calculate half-life. Natural logarithmic models: f(x)=a+bln⁡xf(x)=a+b\ln x. In examinations, students may be expected to interpret and use other models that are introduced in the question.
AHL2.10Scaling very large or small numbers using logarithms. Linearizing data using logarithms to determine if the data has an exponential or a power relationship using best-fit straight lines to determine parameters. Interpretation of log-log and semi-log graphs. In examinations, students will not be expected to draw or sketch these graphs.

Official IB source: Mathematics: applications and interpretation subject brief (PDF, ibo.org).

Quick questions

What is a rational (fractional) exponent?

a1/na^{1/n} means the nnth root of aa, and am/na^{m/n} means that root raised to the power mm: (an)m(\sqrt[n]{a})^m, which is the same as amn\sqrt[n]{a^m}.

How do you solve an exponential equation like a^x = b?

Take logs (usually natural log) of both sides: ln⁡(ax)=ln⁡b\ln(a^x) = \ln b, so xln⁡a=ln⁡bx \ln a = \ln b, giving x=ln⁡bln⁡ax = \dfrac{\ln b}{\ln a}.

What is a natural logarithmic model?

A model of the form f(x)=a+bln⁡xf(x) = a + b\ln x, used when a quantity rises quickly at first and then more slowly, with no upper limit. It is only defined for x>0x > 0.

Why do you take logarithms of data to fit a power or exponential model?

Taking logs turns a curved power or exponential relationship into a straight line, so an ordinary line of best fit can be used to estimate the model's parameters.

How do you simplify a negative rational exponent?

Take the reciprocal of the positive-exponent version: a−m/n=1am/na^{-m/n} = \dfrac{1}{a^{m/n}}. The negative sign flips the fraction, it does not make the value negative.

What if the unknown is both in an exponent and outside it?

Logarithms cannot separate the two places, so graph both sides, for example y=3xy=3^{x} and y=10−xy=10-x, and read the xx-coordinate of each intersection.

How is an exponential model different from a log model?

An exponential model f(x)=kax+cf(x) = ka^x + c has the input in the exponent and is used for growth, or for decay towards the asymptote y=cy=c. A log model f(x)=a+bln⁡xf(x) = a + b\ln x has the input inside the logarithm and models diminishing returns rather than an asymptote.

How do you recover the original parameters after linearising a power model?

For y=axny = ax^n, plotting log⁡y\log y against log⁡x\log x gives a line log⁡y=n(log⁡x)+log⁡a\log y = n(\log x) + \log a. The gradient is nn, and aa is found by raising 10 (or ee) to the power of the yy-intercept.

How do you write a square root as a power?

A square root is the exponent 12\tfrac12, so x=x1/2\sqrt{x}=x^{1/2}. Its reciprocal is 1x=x−1/2\dfrac{1}{\sqrt{x}}=x^{-1/2}.

Does it matter which base of logarithm I use?

No, any base gives the same answer provided you use it on both sides. For 2x=92^{x}=9, x=ln⁡9ln⁡2≈3.17x=\dfrac{\ln9}{\ln2}\approx3.17, and using base 1010 gives the same value.

Ready to practise properly?

Practice questions on every section, marked instantly like the real exam.