IP AMaths Notes (Upper Sec, Year 3-4): 03) Linear Law

Study guideUpdated 06 Nov 2026
Marcus Pang
Reviewed by
Marcus Pang·Managing Director (Maths)

Transform non-linear relationships into straight lines for graph paper work and regression checks in IP AMaths.

For Integrated Programme students: Your current school materials, teacher instructions, and assessment scope take precedence because IP topic sequence and depth vary by school. This is an Eclat IP guide, not the O-Level / SEC G3 exam-track guide.

How this chapter applies

  • Eclat core: transforming power and exponential relationships into straight-line form, reading gradient and intercept, and recovering unknown constants form the main route.
  • School-sensitive extension: reciprocal, mixed, or log-linear models beyond the named K341 families may be school-sensitive depth.
  • 2027 national comparison: K341 Topic G2.5 explicitly requires transforming relationships including y = axn ax^{n} and y = kbx kb^{x} to linear form to determine unknown constants from a straight-line graph.
  • Check your school: confirm permitted axis transformations, plotting precision, and whether regression tools are allowed.
  • Exam-track route: use the separate O-Level and SEC G3 Additional Mathematics notes for K341 topic ownership.
Q: What does IP AMaths Notes (Upper Sec, Year 3-4): 03) Linear Law cover?
A: Transform non-linear relationships into straight lines for graph paper work and regression checks in IP AMaths.

Linear-law questions ask you to recast a model so plotting gives a straight line. Identify the transformation, compute plotting coordinates, and interpret the resulting gradient and intercept.

Keep the full topic roadmap handy via our IP Maths tuition hub so you can jump into related drills, quizzes, or diagnostics as you move through these notes.

New to the Integrated Programme? Start with What is IP? | Browse all free IP notes.

The core idea is simple: Linear Law turns a curved model into a straight-line graph.

Use it as a working check: Decide what to plot on each axis before calculating values. The gradient and intercept must translate back to the original constants.

Then go one layer deeper: Example: for a power model y = A xn x^{n} , plot log(y) against log(x); the gradient gives n and the intercept gives log(A).

Choosing the plotting axes

Before making a table of values, decide what quantity should become the straight-line vertical axis and what should become the horizontal axis. This prevents the common mistake of plotting the original y y against the original x x when the model is not linear.

Model formPlot vertical axisPlot horizontal axisWhat gradient and intercept mean
y=Abx y = Ab^x log⁡y \log y x x Gradient is log⁡b \log b ; intercept is log⁡A \log A .
y=Axn y = Ax^n log⁡y \log y log⁡x \log x Gradient is n n ; intercept is log⁡A \log A .
y=1a+bx y = \dfrac{1}{a + bx} 1y \dfrac{1}{y} x x Gradient is b b ; intercept is a a .
y=a+bln⁡x y = a + b\ln x y y ln⁡x \ln x Gradient is b b ; intercept is a a .
y=Ae−kx y = A\mathrm{e}^{-kx} ln⁡y \ln y x x Gradient is −k -k ; intercept is ln⁡A \ln A .

Common trap: The intercept is often a logged constant, not the constant itself. If the intercept is log⁡A \log A , convert back with A=10intercept A = 10^{\text{intercept}} before stating the model.

Gradient-intercept decoding checkpoint

After drawing the straight-line graph, do not stop at the gradient and intercept. Decode them back into the constants in the original model.

Transformed equation:
  Y = mX + c

Graph tells you:
  gradient = m
  vertical intercept = c

Question usually wants:
  original constants such as A, b, n, a, or k
Original modelStraight-line formWhat the graph givesWhat to report
y=Abxy = Ab^xlog⁡y=xlog⁡b+log⁡A\log y = x\log b + \log Agradient m=log⁡bm = \log b, intercept c=log⁡Ac = \log Ab=10mb = 10^m, A=10cA = 10^c if base-10 logs are used.
y=Axny = Ax^nlog⁡y=nlog⁡x+log⁡A\log y = n\log x + \log Agradient m=nm = n, intercept c=log⁡Ac = \log An=mn = m, A=10cA = 10^c if base-10 logs are used.
y=Ae−kxy = A\mathrm{e}^{-kx}ln⁡y=−kx+ln⁡A\ln y = -kx + \ln Agradient m=−km = -k, intercept c=ln⁡Ac = \ln Ak=−mk = -m, A=ecA = \mathrm{e}^c.

Misconception check: the intercept is not always the original constant. It is the constant in the transformed straight-line equation, so convert it back before writing the final model.

Log-base consistency checkpoint

Once you choose a log base for a transformed graph, keep that base all the way to the final constants. Mixing log⁡\log and ln⁡\ln can change the gradient or intercept values, even when the straight-line idea is the same.

If the graph usesIntercept cc meansConvert back withCommon trap
log⁡10y\log_{10} y against xx for y=Abxy=Ab^xc=log⁡10Ac=\log_{10} AA=10cA=10^cUsing ec\mathrm{e}^c because the model looks exponential.
ln⁡y\ln y against xx for y=Ae−kxy=A\mathrm{e}^{-kx}c=ln⁡Ac=\ln AA=ecA=\mathrm{e}^cUsing 10c10^c after taking natural logs.
log⁡10y\log_{10} y against log⁡10x\log_{10} x for y=Axny=Ax^nc=log⁡10Ac=\log_{10} AA=10cA=10^cReporting cc as AA without undoing the log.

Worked check: if a straight-line graph of ln⁡y\ln y against xx has intercept 1.201.20, then ln⁡A=1.20\ln A=1.20, so A=e1.20≈3.32A=\mathrm{e}^{1.20}\approx3.32. If the graph used log⁡10y\log_{10} y instead, the same intercept value would give A=101.20≈15.8A=10^{1.20}\approx15.8.

Misconception check: the symbol on the graph axis controls the undo step. Do not decide between 10c10^c and ec\mathrm{e}^c from the original word "exponential" alone.

1 Standard transformations

  • Exponential form y=Abx y = Ab^x : take logs to obtain log⁡y=log⁡A+xlog⁡b \log y = \log A + x \log b .
  • Power form y=Axn y = Ax^n : take logs to produce log⁡y=log⁡A+nlog⁡x \log y = \log A + n \log x .
  • Reciprocal form y=1ax+b y = \dfrac{1}{ax + b} : rearrange to 1y=ax+b \dfrac{1}{y} = ax + b .
  • Mixed models (e.g. y=Axn y = \dfrac{A}{x^n} ) combine the above manipulations.

2 Worked example - Power model

Given y=kx1.5 y = kx^{1.5} , rewrite in linear-law form and determine k k from the data (x,y)=(4,32) (x, y) = (4, 32) , (25,500) (25, 500) .

  1. Take base-10 logs: log⁡10y=log⁡10k+1.5log⁡10x \log_{10} y = \log_{10} k + 1.5 \log_{10} x .
  2. Let Y=log⁡10y Y = \log_{10} y , X=log⁡10x X = \log_{10} x , C=log⁡10k C = \log_{10} k .
  3. Compute: X1=log⁡104=0.6021 X_1 = \log_{10} 4 = 0.6021 , Y1=log⁡1032=1.5051 Y_1 = \log_{10} 32 = 1.5051 .
  4. For the second point: X2=log⁡1025=1.3979 X_2 = \log_{10} 25 = 1.3979 , Y2=log⁡10500=2.6990 Y_2 = \log_{10} 500 = 2.6990 .
  5. Gradient of straight line: m=2.6990−1.50511.3979−0.6021=1.500 m = \dfrac{2.6990 - 1.5051}{1.3979 - 0.6021} = 1.500 , matching the exponent in the model.
  6. Intercept: C=Y1−mX1=1.5051−1.500×0.6021=0.60195 C = Y_1 - m X_1 = 1.5051 - 1.500 \times 0.6021 = 0.60195 .
  7. Hence k=100.60195≈4.00 k = 10^{0.60195} \approx 4.00 (3 s.f.).

3 Worked example - Reciprocal model

A dataset follows y=1a+bx y = \dfrac{1}{a + bx} . Show how to plot a straight line and determine a a , b b using the points (1,0.42) (1, 0.42) , (3,0.21) (3, 0.21) .

  1. Rearrange: 1y=a+bx \dfrac{1}{y} = a + bx .
  2. Define Y=1y Y = \dfrac{1}{y} ; plotting Y Y against x x gives intercept a a and gradient b b .
  3. Compute values: for x=1 x = 1 , Y1=10.42=2.381 Y_1 = \dfrac{1}{0.42} = 2.381 ; for x=3 x = 3 , Y2=10.21=4.762 Y_2 = \dfrac{1}{0.21} = 4.762 .
  4. Gradient: b=4.762−2.3813−1=1.1905 b = \dfrac{4.762 - 2.381}{3 - 1} = 1.1905 .
  5. Intercept: a=Y1−b=2.381−1.1905=1.1905 a = Y_1 - b = 2.381 - 1.1905 = 1.1905 .

4 Worked example - Log-linear model

Suppose sensor readings obey y=a+bln⁡x y = a + b \ln x . Given data points (2,4.1) (2, 4.1) , (5,6.7) (5, 6.7) , (9,8.4) (9, 8.4) , estimate a a and b b using a straight-line plot.

  1. Transform the x x -values: compute X=ln⁡x X = \ln x . This gives X1=0.6931 X_1 = 0.6931 , X2=1.6094 X_2 = 1.6094 , X3=2.1972 X_3 = 2.1972 (4 d.p.).
  2. Plot y y against X X ; the model becomes y=a+bX y = a + bX so the gradient equals b b .
  3. Use the first and last points to estimate the gradient: b≈8.4−4.12.1972−0.6931=2.86 b \approx \dfrac{8.4 - 4.1}{2.1972 - 0.6931} = 2.86 .
  4. Find the intercept with any point: a≈4.1−2.86×0.6931=2.12 a \approx 4.1 - 2.86 \times 0.6931 = 2.12 .
  5. State the fitted model: y≈2.12+2.86ln⁡x y \approx 2.12 + 2.86 \ln x . Always mention plotting accuracy limits when quoting constants.

Answer: take a≈2.12 a \approx 2.12 and b≈2.86 b \approx 2.86 .

5 Practice Quiz

Check that you can choose the right straight-line transform, extract gradients/intercepts, and translate them back into model parameters quickly.

6 Try this

Data believed to satisfy y=Ae−kx y = A \mathrm{e}^{-kx} . Outline how you would estimate A A and k k from a semi-log plot, then test with two sample points of your choice.


Sources

  1. National comparator - SEAB - 2027 SEC G3 Additional Mathematics K341 syllabus
  2. MOE - Integrated Programme