Scribe logoScribeNotes & practice
Sign inStart free
  1. Notes
  2. ›IB Mathematics: Applications and Interpretation
  3. ›2 Functions
  4. ›2.10 Scaling and linearizing data

IB Mathematics: Applications and Interpretation · Topic 2 · Functions

2.10 Scaling and linearizing data

Linearise data from power and exponential models, plot the transformed variables, and interpret the gradient and intercept of the resulting line.

  • 4 min read
  • 5 parts
  • 10 questions
  • Draft
  • HL only
Start a Quick 5 on this topic→Jump to the 10 questions
1/5A curve can be straightened by transforming the variables
  1. 1A curve can be straightened by transforming the variables
  2. 2Power models become straight when both variables are logged
  3. 3Exponential models become straight when only the output is logged
  4. 4The gradient and intercept recover the model’s constants
  5. 5The straightest graph shows which model to use
  6. Check your recall
  7. Practice: 10 questions
1

Part 1 of 5: A curve can be straightened by transforming the variables

A straight line is easy to judge by eye, but a curve is not. With a dozen points, you often cannot tell a power curve from an exponential one.

The fix is linearisation. You change the variables you plot, so that data from the right model fall on a straight line. The line’s equation then gives the constants of the model. Two common models are the power model and the exponential model.

The tool is the natural logarithm, written ln⁡\lnln. It turns products into sums and powers into multiples. Two rules do the work: ln⁡(pq)=ln⁡p+ln⁡q\ln(pq)=\ln p+\ln qln(pq)=lnp+lnq and ln⁡(pk)=kln⁡p\ln(p^k)=k\ln pln(pk)=klnp.

Key terms · tap one for its definition

linearisation

Linearisation is the transformation of one or both variables so that data from a non-linear model lie along a straight line.

natural logarithm

The natural logarithm of xxx, written ln⁡x\ln xlnx, is the power to which eee must be raised to give xxx.

Part 1 done

2

Part 2 of 5: Power models become straight when both variables are logged

Doubling the side of a cube multiplies its volume by eight. This is a power model, y=axny=ax^ny=axn, where yyy varies as a power of xxx. The constants aaa and nnn are fixed numbers.

Take natural logarithms of both sides. The rules give ln⁡y=ln⁡a+nln⁡x\ln y=\ln a+n\ln xlny=lna+nlnx. This matches the line Y=mX+cY=mX+cY=mX+c, with Y=ln⁡yY=\ln yY=lny and X=ln⁡xX=\ln xX=lnx. Here mmm is the gradient and ccc is the vertical intercept.

A graph of ln⁡y\ln ylny against ln⁡x\ln xlnx is called a log–log graph. For a power model it is a straight line. Its gradient is nnn and its vertical intercept is ln⁡a\ln alna.

Common mistake

The intercept is ln⁡a\ln alna, not aaa. Reverse the logarithm to recover aaa itself.

Key terms · tap one for its definition

power model

A power model has the form y=axny=ax^ny=axn, where a>0a>0a>0 and nnn are constants and x>0x>0x>0.

log–log graph

A log–log graph plots ln⁡y\ln ylny against ln⁡x\ln xlnx and is a straight line when y=axny=ax^ny=axn.

Part 2 done

Quick check

One question on the part you just read

Easy

State which graph is a straight line when data follow the power model y=axny=ax^ny=axn.
Show answer and explanation

Answer: B · Worked explanation

Taking logarithms gives ln⁡y=ln⁡a+nln⁡x\ln y=\ln a+n\ln xlny=lna+nlnx, which is linear in ln⁡x\ln xlnx. So a graph of ln⁡y\ln ylny against ln⁡x\ln xlnx is straight. A graph of ln⁡y\ln ylny against xxx is the straight form for an exponential model, not for a power model. Checking the form Y=mX+cY=mX+cY=mX+c confirms the match.
3

Part 3 of 5: Exponential models become straight when only the output is logged

Some quantities grow by a fixed factor in equal steps, like a population that doubles every hour. Each step multiplies yyy by the same factor. These follow an exponential model, y=abxy=ab^xy=abx. The same model can be written y=aekxy=ae^{kx}y=aekx.

Take natural logarithms: ln⁡y=ln⁡a+xln⁡b\ln y=\ln a+x\ln blny=lna+xlnb. Here ln⁡y\ln ylny is already linear in xxx, so only yyy needs transforming. A graph of ln⁡y\ln ylny against xxx is a semi-log graph.

Which graph straightens which model

ModelGraph to plotGradientVertical intercept
Power, y=axny=ax^ny=axnln⁡y\ln ylny against ln⁡x\ln xlnxnnnln⁡a\ln alna
Exponential, y=abxy=ab^xy=abxln⁡y\ln ylny against xxxln⁡b\ln blnbln⁡a\ln alna
Exponential, y=aekxy=ae^{kx}y=aekxln⁡y\ln ylny against xxxkkkln⁡a\ln alna
Model
Power, y=axny=ax^ny=axn
Graph to plot
ln⁡y\ln ylny against ln⁡x\ln xlnx
Gradient
nnn
Vertical intercept
ln⁡a\ln alna
Model
Exponential, y=abxy=ab^xy=abx
Graph to plot
ln⁡y\ln ylny against xxx
Gradient
ln⁡b\ln blnb
Vertical intercept
ln⁡a\ln alna
Model
Exponential, y=aekxy=ae^{kx}y=aekx
Graph to plot
ln⁡y\ln ylny against xxx
Gradient
kkk
Vertical intercept
ln⁡a\ln alna

The two graphs differ in what is logged. A power model needs both variables logged. An exponential model needs only yyy logged.

Key terms · tap one for its definition

exponential model

An exponential model has the form y=abxy=ab^xy=abx, or y=aekxy=ae^{kx}y=aekx, where a>0a>0a>0, b>0b>0b>0 and kkk are constants.

semi-log graph

A semi-log graph plots ln⁡y\ln ylny against xxx and is a straight line when y=abxy=ab^xy=abx.

Part 3 done

4

Part 4 of 5: The gradient and intercept recover the model’s constants

Now that each model has a straight-line form, a fitted line gives its constants. Draw the line of best fit through the transformed points. It should pass as close as possible to all of them. Read the gradient from two points that lie far apart on the line.

Wide separation matters, because small reading errors then have less effect. The vertical intercept gives ln⁡a\ln alna. Reverse the logarithm with a=eca=e^{c}a=ec.

Worked example

Fitting a power model

Data follow y=axny=ax^ny=axn with x=1,2,3,4x=1,2,3,4x=1,2,3,4 and y=3.00,12.0,27.0,48.0y=3.00,12.0,27.0,48.0y=3.00,12.0,27.0,48.0. Find aaa and nnn.
  1. 1
    Calculate ln⁡x\ln xlnx: 0, 0.693, 1.099, 1.3860,\ 0.693,\ 1.099,\ 1.3860, 0.693, 1.099, 1.386. Calculate ln⁡y\ln ylny: 1.099, 2.485, 3.296, 3.8711.099,\ 2.485,\ 3.296,\ 3.8711.099, 2.485, 3.296, 3.871.
  2. 2
    The points lie on a straight line, so a power model fits. Use the end points to find the gradient.
  3. 3
    Gradient: n=3.871−1.0991.386−0=2.7721.386=2.00n=\frac{3.871-1.099}{1.386-0}=\frac{2.772}{1.386}=2.00n=1.386−03.871−1.099​=1.3862.772​=2.00.
  4. 4
    The vertical intercept is 1.099, so ln⁡a=1.099\ln a=1.099lna=1.099 and a=e1.099≈3.00a=e^{1.099}\approx3.00a=e1.099≈3.00.

Answer: The model is y=3.00x2.00y=3.00x^{2.00}y=3.00x2.00, that is, y=3x2y=3x^2y=3x2.

Figure 1. After taking logarithms, the four data points lie on one straight line. The line crosses the vertical axis above 1 and rises steadily.

Exam tip

Write the final model in the original variables, such as y=3x2y=3x^2y=3x2. An equation in ln⁡y\ln ylny and ln⁡x\ln xlnx is not the answer.

Part 4 done

Quick check

One question on the part you just read

Easy

State the gradient of the straight line from plotting ln⁡y\ln ylny against xxx when y=abxy=ab^xy=abx.
Show answer and explanation

Answer: D · Worked explanation

Since ln⁡y=ln⁡a+xln⁡b\ln y=\ln a+x\ln blny=lna+xlnb, the coefficient of xxx is ln⁡b\ln blnb, so that is the gradient. The quantity ln⁡a\ln alna is the vertical intercept, which is the tempting wrong choice. The value bbb itself is found afterwards by reversing the logarithm. Always finish by reversing the logarithm when the model constant itself is needed.
5

Part 5 of 5: The straightest graph shows which model to use

Often the model is not known in advance. Try each graph and see which one gives a straight line. If none does, none of these models suits the data.

  • yyy against xxx is straight: a linear model.
  • ln⁡y\ln ylny against ln⁡x\ln xlnx is straight: a power model.
  • ln⁡y\ln ylny against xxx is straight: an exponential model.

Worked example

Fitting an exponential model

The number of bacteria NNN in a culture is measured at time ttt hours: (0,150)(0,150)(0,150), (2,273)(2,273)(2,273), (4,498)(4,498)(4,498), (6,907)(6,907)(6,907). Find aaa and kkk in N=aektN=ae^{kt}N=aekt.
  1. 1
    Take logarithms: ln⁡N=ln⁡a+kt\ln N=\ln a+ktlnN=lna+kt. Plot ln⁡N\ln NlnN against ttt.
  2. 2
    Calculate ln⁡N\ln NlnN: 5.011, 5.609, 6.211, 6.8105.011,\ 5.609,\ 6.211,\ 6.8105.011, 5.609, 6.211, 6.810. These rise by almost equal steps, so the graph is straight.
  3. 3
    Gradient: k=6.810−5.0116−0=0.300k=\frac{6.810-5.011}{6-0}=0.300k=6−06.810−5.011​=0.300, in h−1\mathrm{h^{-1}}h−1.
  4. 4
    The intercept at t=0t=0t=0 is 5.011, so a=e5.011≈150a=e^{5.011}\approx150a=e5.011≈150.

Answer: The model is N≈150e0.300tN\approx150e^{0.300t}N≈150e0.300t, with ttt in hours.

With a model found, you can predict values. Predictions inside the range of the data are more reliable than predictions outside it.

Common mistake

The gradient of the straight line is not the gradient of the original curve. For ln⁡y\ln ylny against xxx, it is the constant kkk in y=aekxy=ae^{kx}y=aekx.

Go deeperWhy the base of the logarithm hardly matters

Common logarithms, written log⁡\loglog, also straighten both models. For a power model, using log⁡\loglog on both axes leaves the gradient equal to nnn. The intercept becomes log⁡a\log aloga.

For an exponential model, the gradient becomes log⁡b\log blogb instead of ln⁡b\ln blnb. Either way, reversing the same logarithm recovers bbb.

Note read

Check your recall[15 marks]

6 written questions. Answer each one in your head or on paper, then open the mark scheme.

  1. 1.State[3 marks]

    State which graph linearises data that follow y=axny=ax^ny=axn, and state what its gradient and vertical intercept represent.

    Show mark schemeHide mark scheme

    Model answer

    Plot ln⁡y\ln ylny against ln⁡x\ln xlnx. The gradient is nnn and the vertical intercept is ln⁡a\ln alna.

    • 1 mark for stating that ln⁡y\ln ylny is plotted against ln⁡x\ln xlnx
    • 1 mark for stating that the gradient is nnn
    • 1 mark for stating that the vertical intercept is ln⁡a\ln alna
  2. 2.Explain[3 marks]

    Explain why a graph of ln⁡y\ln ylny against xxx is a straight line when y=abxy=ab^xy=abx.

    Show mark schemeHide mark scheme

    Model answer

    Taking logarithms gives ln⁡y=ln⁡a+xln⁡b\ln y=\ln a+x\ln blny=lna+xlnb. This has the form Y=mX+cY=mX+cY=mX+c with Y=ln⁡yY=\ln yY=lny and X=xX=xX=x. The gradient ln⁡b\ln blnb and intercept ln⁡a\ln alna are constants, so the graph is straight.

    • 1 mark for taking logarithms to obtain ln⁡y=ln⁡a+xln⁡b\ln y=\ln a+x\ln blny=lna+xlnb
    • 1 mark for matching this to the straight-line form Y=mX+cY=mX+cY=mX+c
    • 1 mark for stating that ln⁡b\ln blnb and ln⁡a\ln alna are constants
  3. 3.Determine[3 marks]

    A graph of ln⁡y\ln ylny against xxx is a straight line with gradient 0.250 and vertical intercept 2.00. Determine aaa and kkk in the model y=aekxy=ae^{kx}y=aekx.

    Show mark schemeHide mark scheme

    Model answer

    The gradient gives k=0.250k=0.250k=0.250. The intercept gives ln⁡a=2.00\ln a=2.00lna=2.00, so a=e2.00≈7.39a=e^{2.00}\approx7.39a=e2.00≈7.39.

    • 1 mark for identifying kkk as the gradient, 0.250
    • 1 mark for using a=e2.00a=e^{2.00}a=e2.00 from the intercept
    • 1 mark for obtaining a≈7.39a\approx7.39a≈7.39
  4. 4.Calculate[2 marks]

    A model is y=7.39e0.250xy=7.39e^{0.250x}y=7.39e0.250x. Calculate yyy when x=4.00x=4.00x=4.00.

    Show mark schemeHide mark scheme

    Model answer

    y=7.39e0.250×4.00=7.39e1.00≈20.1y=7.39e^{0.250\times4.00}=7.39e^{1.00}\approx20.1y=7.39e0.250×4.00=7.39e1.00≈20.1.

    • 1 mark for substituting x=4.00x=4.00x=4.00 to obtain 7.39e1.007.39e^{1.00}7.39e1.00
    • 1 mark for obtaining y≈20.1y\approx20.1y≈20.1
  5. 5.Distinguish[2 marks]

    Distinguish between a log–log graph and a semi-log graph.

    Show mark schemeHide mark scheme

    Model answer

    A log–log graph plots ln⁡y\ln ylny against ln⁡x\ln xlnx and suits a power model. A semi-log graph plots ln⁡y\ln ylny against xxx and suits an exponential model.

    • 1 mark for stating that a log–log graph logs both variables and suits a power model
    • 1 mark for stating that a semi-log graph logs only yyy and suits an exponential model
  6. 6.Deduce[2 marks]

    Deduce which model fits data when ln⁡y\ln ylny against ln⁡x\ln xlnx is a straight line but ln⁡y\ln ylny against xxx is curved.

    Show mark schemeHide mark scheme

    Model answer

    A power model fits, because a straight log–log graph shows ln⁡y\ln ylny is linear in ln⁡x\ln xlnx.

    • 1 mark for deducing a power model
    • 1 mark for linking this to ln⁡y\ln ylny being linear in ln⁡x\ln xlnx

Practice this topic

10 quick-fire multiple-choice questions, easy to hard. Tap an answer to see if you're right and the worked explanation. (The written questions with mark schemes are under “Check your recall” above.)

0 right0 of 10 answered

Includes the 2 quick checks in the note above.

3.

Easy[1 mark]

Which statement about natural logarithms is correct?
Show answer and explanation

Answer: B · Worked explanation

A power inside a logarithm becomes a multiplier, so ln⁡(pk)=kln⁡p\ln(p^k)=k\ln pln(pk)=klnp. This rule turns xnx^nxn into nln⁡xn\ln xnlnx. The expression (ln⁡p)k(\ln p)^k(lnp)k raises the logarithm itself to a power, which is a different quantity. Also, ln⁡(pq)\ln(pq)ln(pq) equals a sum of logarithms, not a product.
4.

Medium·Determine[1 mark]

Data follow y=ax3y=ax^3y=ax3. A graph of ln⁡y\ln ylny against ln⁡x\ln xlnx is a straight line with vertical intercept 2.00. Determine aaa.
Show answer and explanation

Answer: B · Worked explanation

The vertical intercept equals ln⁡a\ln alna, so ln⁡a=2.00\ln a=2.00lna=2.00 and a=e2.00≈7.39a=e^{2.00}\approx7.39a=e2.00≈7.39. The value 2.00 mistakes the intercept for aaa itself. The gradient of the line would be 3, the power, and that value does not give aaa. Remember that the line of ln⁡y\ln ylny against ln⁡x\ln xlnx has a gradient equal to the power nnn, here 3.
5.

Medium·Determine[2 marks]

A graph of ln⁡y\ln ylny against ttt is a straight line through (0,4.00)(0,4.00)(0,4.00) and (5,6.00)(5,6.00)(5,6.00). For the model y=aekty=ae^{kt}y=aekt, determine kkk.
Show answer and explanation

Answer: A · Worked explanation

The gradient is 6.00−4.005−0=0.400\frac{6.00-4.00}{5-0}=0.4005−06.00−4.00​=0.400, and for y=aekty=ae^{kt}y=aekt this gradient is kkk. The value 4.00 is the intercept, which equals ln⁡a\ln alna. The value 2.50 comes from inverting the fraction, dividing the horizontal change by the vertical change. Always divide the vertical change by the horizontal change.
6.

Medium·Determine[2 marks]

A quantity follows y=axny=ax^ny=axn. When x=2x=2x=2, y=12y=12y=12. When x=4x=4x=4, y=48y=48y=48. Determine nnn.
Show answer and explanation

Answer: A · Worked explanation

Doubling xxx from 2 to 4 multiplies yyy by 4812=4\frac{48}{12}=41248​=4, so 2n=42^n=42n=4 and n=2n=2n=2. Equivalently, the gradient is ln⁡48−ln⁡12ln⁡4−ln⁡2=ln⁡4ln⁡2=2\frac{\ln48-\ln12}{\ln4-\ln2}=\frac{\ln4}{\ln2}=2ln4−ln2ln48−ln12​=ln2ln4​=2. The value 4 is the factor by which yyy changes, not the power. As a check, 22=42^2=422=4 matches the observed factor, whereas 23=82^3=823=8 does not.
7.

Medium·Deduce[1 mark]

Deduce which model fits data when a graph of ln⁡y\ln ylny against xxx is straight but a graph of ln⁡y\ln ylny against ln⁡x\ln xlnx is curved.
Show answer and explanation

Answer: A · Worked explanation

A straight line of ln⁡y\ln ylny against xxx matches ln⁡y=ln⁡a+xln⁡b\ln y=\ln a+x\ln blny=lna+xlnb, the form of an exponential model. A power model would give a straight line when both variables are logged, which these data do not show. The gradient of that line would equal ln⁡b\ln blnb.
8.

Hard·Determine[2 marks]

Data follow y=abxy=ab^xy=abx. A straight-line graph of ln⁡y\ln ylny against xxx passes through (1,2.30)(1,2.30)(1,2.30) and (3,4.60)(3,4.60)(3,4.60). Determine bbb.
Show answer and explanation

Answer: A · Worked explanation

The gradient is 4.60−2.303−1=1.15\frac{4.60-2.30}{3-1}=1.153−14.60−2.30​=1.15, and this equals ln⁡b\ln blnb. So b=e1.15≈3.16b=e^{1.15}\approx3.16b=e1.15≈3.16. The value 1.15 is the gradient itself, not bbb, which is the tempting error. The logarithm must be reversed. The intercept is also ln⁡a=1.15\ln a=1.15lna=1.15, so a≈3.16a\approx3.16a≈3.16 as well.
9.

Hard·Predict[2 marks]

A semi-log graph has the straight-line equation ln⁡y=0.500x+1.00\ln y=0.500x+1.00lny=0.500x+1.00, with y=aekxy=ae^{kx}y=aekx. Predict yyy when x=6x=6x=6.
Show answer and explanation

Answer: A · Worked explanation

Substitute x=6x=6x=6: ln⁡y=0.500×6+1.00=4.00\ln y=0.500\times6+1.00=4.00lny=0.500×6+1.00=4.00. Reversing the logarithm gives y=e4.00≈54.6y=e^{4.00}\approx54.6y=e4.00≈54.6. Stopping at 4.00 reports ln⁡y\ln ylny, not yyy. The value e3≈20.1e^{3}\approx20.1e3≈20.1 ignores the intercept, which equals ln⁡a\ln alna. Always check which quantity the equation gives: the left side is a logarithm, so one more step is needed.
10.

Hard·State[2 marks]

The data follow y=5x2y=5x^2y=5x2. State the gradient and vertical intercept of the graph of ln⁡y\ln ylny against ln⁡x\ln xlnx.
Show answer and explanation

Answer: A · Worked explanation

Taking logarithms gives ln⁡y=2ln⁡x+ln⁡5\ln y=2\ln x+\ln5lny=2lnx+ln5. The gradient is the power, 2, and the vertical intercept is ln⁡5≈1.61\ln5\approx1.61ln5≈1.61. Gradient 5 with intercept 2 swaps the roles of the two constants. Quoting 5 as the intercept forgets that ln⁡a\ln alna appears, not aaa. Reversing the logarithms of ln⁡y=2ln⁡x+ln⁡5\ln y=2\ln x+\ln5lny=2lnx+ln5 gives y=5x2y=5x^2y=5x2, which confirms the match.

Keep going

Next topic3.1 3D distance, volume and surface areaPrevious topic2.9 Further models

More in Topic 2 · Functions

  1. 2.2Functions, domain and rangeNotes · 10 questions
  2. 2.3Graphs of functionsNotes · 10 questions
  3. 2.4Key features of graphsNotes · 10 questions
  4. 2.5Modelling with functionsNotes · 10 questions
  5. 2.6The modelling processNotes · 10 questions
  6. 2.7Composite and inverse functionsNotes · 10 questions· HL only
  7. 2.8Transformations of graphsNotes · 10 questions· HL only
  8. 2.9Further modelsNotes · 10 questions· HL only
  9. 2.10Scaling and linearizing dataNotes · 10 questions· HL only

All IB Mathematics: Applications and Interpretation topics →

Draft: written by Scribe and checked automatically, not yet by a subject reviewer.

Practice questions are original, written by Scribe and checked automatically for correctness.

Topic codes follow the official numbering. Not endorsed by IB Diploma Programme or any exam board.

On this page

0/5 read

  1. 1A curve can be straightened by transforming the variables
  2. 2Power models become straight when both variables are logged
  3. 3Exponential models become straight when only the output is logged
  4. 4The gradient and intercept recover the model’s constants
  5. 5The straightest graph shows which model to use
  6. Check your recall
  7. Practice: 10 questions
All coursesScribeHow it worksPricingPrivacyTerms

Notes and questions are written by Scribe in our own words. Topic codes follow the official course numbering so you can find your place. Not produced or endorsed by the IB, College Board, AQA, OCR, Pearson Edexcel or any exam board.

© 2026 Scribe

Straightened power data

ln y (unitless)00.20.40.60.811.21.41.601234ln x (unitless)
  • Line of best fit
  • Data