MOE V3.0 Ref 320 | A-Math Linear Law and Log Graphs

Did you know that a curve can become a straight line without changing the underlying relationship? For many Secondary G3 Additional Mathematics models, the trick is to transform the plotted quantities—not to pretend the original data are linear. The 2027 syllabus specifically includes recovering constants from relationships such as y=axⁿ and y=kbˣ when drawn on the appropriate logarithmic axes.

This Ministry of Education V3.0 reference teaches K341 coordinate geometry G2 linearisation through the difference between power models and exponential models, correct axes, slope, intercept, inverse logarithms and original-data checks. It provides twelve stages, twenty-six original worked examples, student transfer tasks, a misconception ledger and eduKate Depth–Load–Transfer diagnostics.

This is independent eduKateSG teaching material, not an official examination paper or claims about actual engineering datasets. All plotted points and contexts below are invented, exact mathematical examples unless described as approximations.

1. A Curve Can Tell a Straight-Line Story

The plotted relation y=2x³ is curved on ordinary x–y axes. For positive x and y, logarithms transform it into ln y=3ln x+ln2, which is a straight line when the horizontal axis is ln x and vertical axis is ln y. The graph is a different representation of the same numbers. A student’s duty is to identify what the axes actually represent.

2. Exact 2027 G3 Additional Mathematics Scope

The 2027 SEAB G3 Additional Mathematics K341 syllabus, page 7, Geometry and Trigonometry G2 includes the transformation of given relationships, specifically y=axⁿ and y=kbˣ, to linear form for determining unknown constants from a straight-line graph. It appears under coordinate geometry in two dimensions, alongside midpoints, line conditions, rectilinear area and circle equations. The use of logarithms draws on A6, but this chapter focuses on extracting constants from transformed straight-line plots.

3. Two Essential Transformations

Original relationship	Horizontal coordinate X	Vertical coordinate Y	Line slope	Intercept
y=a xⁿ, with a>0 and x>0	ln x	ln y	n	ln a
y=k bˣ, with k>0 and b>0	x	ln y	ln b	ln k
y=a xⁿ, common-log version	log₁₀ x	log₁₀ y	n	log₁₀ a
y=k bˣ, common-log version	x	log₁₀ y	log₁₀ b	log₁₀ k

Both axes must be read literally. The first case plots two logged variables; the second plots one logged variable against original x. Recover coefficients by exponentiating an intercept or slope as appropriate; do not leave an answer as the logarithm of the required constant.

4. Eight Rooms of Transformational Reasoning

Room one recognises the model family. Room two checks the positive domain. Room three selects axis variables. Room four takes consistent logarithms. Room five finds straight-line slope. Room six recovers constants from slope and intercept. Room seven checks original data and context. Room eight transfers the technique to unfamiliar plotted coordinates.

5. Twelve-Stage Log–Plot–Recover Protocol

Stage 1 — Recognise the original relationship

Decide whether a model resembles a power law y=axⁿ or an exponential law y=kbˣ. Their variables occupy different positions; x is an exponent in one and a base being raised to a fixed power in the other.

Stage 2 — Write domain restrictions before taking logs

For a log-log plot of y=axⁿ, x>0 and y>0 are needed, with a>0 in the usual model. For log y against x using y=kbˣ, k>0 and b>0 ensure positive y. Never silently log zero or a negative sample.

Stage 3 — Choose which quantities form axes

If Y=ln y and X=ln x, a power law turns into Y=nX+ln a. If Y=ln y and X=x, an exponential law turns into Y=(ln b)X+ln k. The horizontal axis differs.

Stage 4 — Take the same logarithm base consistently

ln y and ln x use natural logs; log₁₀ y and log₁₀ x use common logs. If different bases are used across axes, the slope may acquire an extra conversion factor; do not claim it always equals n.

Stage 5 — Match the equation with Y=mX+c

Read off slope m and vertical intercept c on the transformed graph. Label whether the printed graph shows x, ln x, y, ln y or log₁₀ quantities. Never substitute the intercept as a raw a without reversing its logarithm.

Stage 6 — Interpret slope physically or mathematically

For a power relationship, slope n on a log-log graph is the power-law exponent. For an exponential law with natural log y against x, slope ln b, so b=eᵐ. The sign indicates increasing or decreasing positive models.

Stage 7 — Recover the multiplier from the intercept

For ln transformation, a=eᶜ or k=eᶜ. For common log transformation, a=10ᶜ or k=10ᶜ. The intercept is a logarithm of the coefficient, not usually the coefficient itself.

Stage 8 — Use two reliable transformed points

Compute m=(Y₂−Y₁)/(X₂−X₁), avoiding equal X values. Then substitute into Y=mX+c to find the intercept. This is the same coordinate-geometry logic as ordinary straight lines.

Stage 9 — Test a third point and the original data

Using x values in the original model should give predicted y values consistent with those supplied or with the specified plotting precision. A line fitted to noisy data may only approximately represent a model.

Stage 10 — Keep axis scaling and units explicit

On a graph printed as ln y against ln x, a unit on the axis represents a logarithmic quantity. If physical units are attached to x and y, the correct use of reference units can be necessary before logarithms; school questions usually define numerical graph axes.

Stage 11 — Interpret model coefficients in context

A positive n means y rises with x on x>0; a negative n means an inverse-power relationship. If 0

Stage 12 — Verify by reversing the transformation

Exponentiate the transformed line equation or reconstruct y=a xⁿ / y=k bˣ. Substitute a small sample, check the graph axes once more, and reject results violating domain or meaningful context.

6. Original Fictional Case: A Growth and Scale Investigation

Two imaginary school groups collect positive numerical data. Group A measures a quantity that follows a power y=axⁿ when size x changes. Group B investigates a quantity that multiplies by the same factor at successive equal x steps, suggesting y=kbˣ. Both begin with curved-looking original graphs. Their challenge is to choose the transformation, derive the constants and show at least one original value that their model predicts correctly.

7. Twenty-Six Original Worked Linear-Law Problems

8. Worked Example A — A cubic power relationship

Given y=2x³ for x>0, natural logs give ln y=3ln x+ln 2. On axes X=ln x, Y=ln y, the slope is 3 and intercept ln 2. The slope is not 2; 2 is the original multiplier.

Teaching audit: record the axes, whether natural or common logarithms were used, the meaning of slope and intercept, and a check on the original unlogged relationship. A line’s numerical slope has no reliable meaning until its axis titles are known.

9. Worked Example B — Reconstruct a power law from a line

A plot of Y=ln y against X=ln x gives Y=2X+ln 5. Therefore ln y=2ln x+ln 5=ln(5x²), so y=5x² for positive x. The y-intercept is ln 5, not 5.

10. Worked Example C — Negative power on log-log axes

If Y=ln y and X=ln x lie on Y=−2X+ln 10, the original relation is y=10/x² for x>0. The transformed slope is negative because larger x gives smaller y.

11. Worked Example D — Power law with exponent one-half

For y=3√x on positive x, ln y=0.5ln x+ln 3. The log-log line slope is 1/2 and its natural-log intercept corresponds to multiplier 3.

12. Worked Example E — Same power law using base-ten logs

For y=4x², log₁₀y=2log₁₀x+log₁₀4. Both log bases yield slope 2 when used consistently, though their plotted intercept numbers differ.

13. Worked Example F — A common-log intercept of two

A line Y=−3X+2 with Y=log₁₀y, X=log₁₀x gives y=10²x⁻³=100/x³. Coefficient is 100, not 2.

14. Worked Example G — Slope from two transformed pairs

A log-log line goes through X=0,Y=ln 3 and X=ln 2,Y=ln 12. Slope=(ln12−ln3)/ln2=ln4/ln2=2. Intercept ln3, giving y=3x².

15. Worked Example H — Power formula checked in the original scale

With y=3x² and x=4, y=48. Taking logs yields ln48=ln3+2ln4. Reconstructing a numerical datum provides an independent reasonableness check.

16. Worked Example I — The x=1 anchor for a power law

If y=axⁿ and x=1, then y=a because 1ⁿ=1. On the log-log graph, X=ln1=0 and Y=ln a, which explains the transformed intercept.

17. Worked Example J — The x=0 impossibility on a log-log graph

A mathematical model might define y=x² at x=0, but ln x is undefined there. The log-log graph cannot include X=ln0. Domain restrictions belong to the transformation even if an original function has a wider domain.

18. Worked Example K — A doubling rule

For y=6x³, doubling x multiplies y by 2³=8. The slope 3 of its log-log line encodes this exponent; this is not the same as an exponential doubling with every one-unit increase in x.

19. Worked Example L — An exponential base of two

For y=5×2ˣ, ln y=(ln2)x+ln5. Plot ln y against the original x, not ln x. Slope ln2; intercept ln5.

20. Worked Example M — Reconstruct an exponential model

A line Y=0.7X+ln4 with Y=ln y and X=x gives y=4e^(0.7x). This is k bˣ with k=4 and b=e^0.7; 0.7 itself is not the base.

21. Worked Example N — An exact exponential base from a slope

If ln y plotted against x has slope ln3 and intercept ln2, then y=2×3ˣ. Exponentiating both slope and intercept recovers base 3 and multiplier 2.

22. Worked Example O — Exponential decline

For y=100(0.8)ˣ, ln y=ln100+x ln0.8. Since ln0.8<0, the semi-log line has negative slope. The output remains positive for all real x.

23. Worked Example P — Common-log semi-log line

For y=7×10^(2x), log₁₀y=2x+log₁₀7. On axes log₁₀ y against x the slope is 2, so the underlying growth base is 10²=100 per unit x.

24. Worked Example Q — A base-ten semi-log slope of 0.3010

If Y=log₁₀y plotted against x has slope approximately 0.3010 and intercept 1, then b≈10^0.3010≈2 and k=10, so y≈10×2ˣ. Numerical rounding makes the model approximate.

25. Worked Example R — Use two exponential data points

An exponential model y=kbˣ has values y=6 at x=0 and y=18 at x=1. Then k=6 and b=18/6=3. Therefore y=6×3ˣ and the semi-log slope is ln3.

26. Worked Example S — Use points separated by two in x

For y=kbˣ with y(0)=2 and y(2)=18, k=2 and b²=9, so b=3 because the model takes a positive base. The relation is y=2×3ˣ.

27. Worked Example T — A negative exponential slope

If Y=ln y against x has slope −ln2 and intercept ln8, then y=8×(1/2)ˣ. The base is e^(−ln2)=1/2, not −2.

28. Worked Example U — Read the right axes

A line has slope 2 on axes ln y against ln x: that means n=2 in a power law. The same numerical slope on axes ln y against x means the exponential base is e². Axis titles are essential to interpretation.

29. Worked Example V — Do not mix logarithm bases

If Y=ln y is plotted against X=log₁₀x for a power y=axⁿ, then ln x=(ln10)X and slope is n ln10. Reading the slope directly as n would be incorrect.

30. Worked Example W — An unphysical negative coefficient

A student proposes ln y=ln(−3)+2ln x on a real-valued log-log graph. ln(−3) is undefined in reals, so the standard transformed equation is invalid. The original model’s domain and coefficient assumptions must be checked.

31. Worked Example X — A non-linear transformed plot

If the data for ln y against ln x trace a clear curve instead of a straight line, the simple y=axⁿ model may be unsuitable across the measured range. Recheck axes, data and the model rather than forcing a power law.

32. Worked Example Y — Rounding a graph intercept

Suppose a plotted common-log intercept is about 0.477. Then a≈10^0.477≈3.0, not exactly 0.477. Include appropriate significant figures because a read-off graph rarely determines an infinitely precise constant.

33. Worked Example Z — A fully verified power/exponential comparison

The table x=1,2,4 with y=3,12,48 fits y=3x²; ratios of y are 4 and 4 for doubled x. An exponential y=kbˣ with a fixed b per one-unit x step does not fit all those points. Use transformed axes and the original table to decide the family.

34. A straight line appears only on the correct transformed axes

For a power law use log y against log x. For an exponential law use log y against x. The variables defined for X and Y determine what slope and intercept actually mean.

35. Exponentiating the intercept is essential

The intercept ln a or log₁₀a is not the original multiplier a. Recover the coefficient with e^c or 10^c according to the log base.

36. The distinction between x as a base and x as an exponent is fundamental

In y=axⁿ the variable x is the base under a fixed exponent; in y=kbˣ the variable x appears in the exponent. Their trends and transformed plots are different.

37. Consistent log bases preserve the power exponent slope

If both axes use the same log base, the log-log slope remains n. Using a mixed pair of bases changes the scale and can change the numerical slope.

38. Only meaningful data should be transformed

Logarithms of non-positive real values are undefined. Experimental data may contain zeros or signs that require a different model or careful treatment rather than blind calculation.

39. Two points determine a theoretical line but not truth of a model

Any two distinct transformed points define a line. A third point or wider data range tests whether the supposed relationship remains plausible.

40. Graph reading involves approximation

A drawn slope and intercept may have limited precision. Students should report appropriately rounded model parameters and consider whether discrepancies are within plotting accuracy.

41. A power exponent can be negative

A negative slope on a log-log plot corresponds to y decreasing with x for a positive x domain. A negative exponent does not mean y must be negative.

42. An exponential decay base lies between zero and one

A negative slope on ln y against x implies 00. The output can decline while remaining positive.

43. Modelling is not an assertion of causality

A log-linear pattern describes a relationship between variables. It does not prove that changing one physically causes a change in the other without further evidence.

44. Correct units and graph labels protect conclusions

When a problem prints ln(y/y₀) or other dimensionless ratios, retain their meaning. The logarithm of a unit-bearing physical quantity needs careful interpretation beyond simplified numeric school graphs.

45. A model should always return to the original scale

After finding a and n or k and b, predict y for at least one x value and compare against the supplied table. This catches slope/base and intercept mistakes quickly.

46. Misconception Ledger

Treating a power law and an exponential law as identical models.
Taking ln y but forgetting whether the horizontal axis is x or ln x.
Reading log intercept c directly as a or k instead of taking eᶜ or 10ᶜ.
Using slope n for a semi-log exponential graph instead of recovering b=eᵐ or 10ᵐ.
Mixing ln and log₁₀ on the two axes without accounting for scale conversion.
Taking logarithms of non-positive values.
Assuming any two transformed points prove the model is valid across all data.
Failing to test fitted constants in the original relationship.
Treating a negative exponent as automatically negative y.
Quoting graph-read constants to impossible precision.
Claiming a straight-line association proves physical causation.
Copying a y-intercept from the original graph when the question gives a transformed graph.

47. Three-Pupil Linearisation Workshop

Pupil A identifies whether each original dataset is better tested with log y versus log x or log y versus x. Pupil B finds slope and intercept from clearly marked transformed points. Pupil C recovers a,n or k,b and checks original readings. Rotate roles and then give each learner an unfamiliar graph with the axis labels hidden initially; ask what information must be known before deciding the constants.

48. eduKate Depth–Load–Transfer Diagnostic

Depth: derive both transformed equations without referring to a template. Load: read two transformed points, compute the gradient and intercept, recover the constants, and check a third original data point. Transfer: change from a growth power relationship to exponential decay and switch from natural to common logarithms. Tag mistakes as family, log base, domain, slope, intercept, inverse transformation or model validation, then repair the earliest weak room.

49. Independent Retrieval and Checked Answers

Try these: (1) ln y=4 ln x+ln 3 gives y=3x⁴. (2) log₁₀ y=−2 log₁₀ x+1 gives y=10/x². (3) ln y=(ln 2)x+ln 5 gives y=5×2ˣ. (4) ln y=−(ln 3)x+ln 9 gives y=9×(1/3)ˣ. (5) x=1 gives y=a for power y=axⁿ. (6) x=0 gives y=k for exponential y=kbˣ. Each response should state whether X=ln x or X=x, not merely the final formula.

50. Parent and Tutor Questions

Ask ‘What exactly is on each graph axis?’, ‘Why are the logs defined here?’, ‘What does the slope mean in this model?’, ‘Is the intercept a number or a logarithm of a constant?’, ‘Could the graph be exponential instead?’, ‘What happens when the x value doubles?’, and ‘Does your reconstructed equation predict the original data?’ These questions make a transform a reasoning tool instead of a chant.

51. Frequently Asked Questions

What is linearisation?

Rewriting a curved algebraic relationship so suitable transformed axes produce a straight-line form Y=mX+c, which helps recover unknown model constants.

Is linear law still in the 2027 SEC G3 Additional Mathematics syllabus?

Yes. K341 G2 explicitly includes transforming relationships y=axⁿ and y=kbˣ to straight-line form to determine unknown constants.

For y=axⁿ, which axes should be used?

Logarithm of y against logarithm of x using a consistent base for positive values. The slope is n and the intercept is log a.

For y=kbˣ, which axes are used?

Logarithm of y against the original x. The slope is log b and the intercept log k.

If the y-intercept is 2, is a=2?

Not on log y axes. If the graph uses ln y, a or k=e². If it uses log₁₀y, a or k=10².

Can a logarithmic graph include x=0?

Not when the horizontal axis is log x; log 0 is undefined. In an exponential semi-log plot with x unlogged, x=0 is possible if y>0.

Does a straight graph prove causation?

No. It supports compatibility with a chosen mathematical model across the observed range, not the physical cause of the relationship.

Does the task require logarithmic regression software?

The K341 requirement is to transform given relationships and determine constants from a straight-line graph; advanced statistical regression software is not specified as a compulsory subtopic.

52. Closing Principle

Start with the original family, transform the correct coordinates, recover coefficients by reversing the logarithms and check against original points. The straight line is a tool for understanding a curved model, not a new mathematical law or a substitute for checking assumptions.

Related eduKateSG References and Official Sources

Read Reference 282 — Coordinate Geometry of Circles, Reference 257 — Exponential and Logarithmic Functions, Reference 184 — Coordinates and Linear Graphs, MOE V3.0 library, the immutable Clementi teaching benchmark and official SEAB 2027 K341 syllabus, page 7, G2.

MOE V3.0 Reference Sequence Navigation

Previous reference 319 · MOE V3.0 reference library. Transfer the transformation to a new graph using different logarithm axes.

Ministry of Education V3.0 Reference 320 | A-Math Linear Law: Transform the Graph, Recover the Constants