Scribe logoScribeNotes & practice
Sign inStart free
  1. Notes
  2. ›IB Mathematics: Applications and Interpretation
  3. ›2 Functions
  4. ›2.9 Further models

IB Mathematics: Applications and Interpretation · Topic 2 · Functions

2.9 Further models

Select a suitable function for a real situation, compare different models and describe the limitations of a chosen model.

  • 5 min read
  • 6 parts
  • 10 questions
  • Draft
  • HL only
Start a Quick 5 on this topic→Jump to the 10 questions
1/6Different situations call for different families of function
  1. 1Different situations call for different families of function
  2. 2The pattern in the data suggests a suitable model
  3. 3Key features of the data give the parameters
  4. 4A logistic model levels off at a limit
  5. 5Comparing models means judging both fit and sense
  6. 6Every model has limitations that should be stated
  7. Check your recall
  8. Practice: 10 questions
1

Part 1 of 6: Different situations call for different families of function

Real situations are rarely a straight line. A wider set of functions lets you model more of them. Each family has a typical shape and a typical use.

A sinusoidal function repeats a smooth wave, so it suits tides and daylight hours. A logistic function grows in an S-shape and levels off, so it suits a population with limited resources.

A piecewise function uses different rules on different intervals, such as a delivery price that changes above a set weight.

Common model families

ModelKey featureExample context
LinearConstant rate of changeFixed fee plus a rate
QuadraticOne turning pointPath of a thrown ball
ExponentialConstant percentage changeCompound interest
SinusoidalRepeating waveTide height
LogisticS-shape that levels offPopulation with limited food
PiecewiseDifferent rule per intervalTiered prices
Model
Linear
Key feature
Constant rate of change
Example context
Fixed fee plus a rate
Model
Quadratic
Key feature
One turning point
Example context
Path of a thrown ball
Model
Exponential
Key feature
Constant percentage change
Example context
Compound interest
Model
Sinusoidal
Key feature
Repeating wave
Example context
Tide height
Model
Logistic
Key feature
S-shape that levels off
Example context
Population with limited food
Model
Piecewise
Key feature
Different rule per interval
Example context
Tiered prices

Key terms · tap one for its definition

sinusoidal function

A sinusoidal function is a function whose graph is a smooth wave that repeats at regular intervals.

logistic function

A logistic function is an S-shaped function of the form L1+Ce−kx\frac{L}{1+Ce^{-kx}}1+Ce−kxL​ that increases towards an upper value LLL.

piecewise function

A piecewise function is a function defined by different rules on different intervals of its domain.

Part 1 done

2

Part 2 of 6: The pattern in the data suggests a suitable model

With these families in mind, the next task is choosing one. Start by plotting the data and describing its shape. A steady rise suggests a linear model. A repeating wave suggests a sinusoidal model.

The context matters as much as the shape. A tide does not grow without limit, and a population cannot be negative. The model must make sense for the quantity being measured.

A graphing calculator can fit several families to the same data. It gives you the parameters, but you must still decide which family is sensible.

Exam tip

Justify your choice using both the pattern in the data and the context of the problem.

Part 2 done

Quick check

One question on the part you just read

Easy

Which type of function is best suited to the height of a tide over several days?
Show answer and explanation

Answer: B · Worked explanation

A tide rises and falls in a regular repeating pattern, so a sinusoidal function suits it. An exponential function grows or decays without repeating. A logistic function levels off but does not oscillate. A linear function changes at a constant rate.
3

Part 3 of 6: Key features of the data give the parameters

After choosing a family, you must find its parameters. A sinusoidal model has the form h=asin⁡(bt)+dh=a\sin(bt)+dh=asin(bt)+d, which starts on the midline and rises. Its amplitude is ∣a∣|a|∣a∣, and ddd is the midline value, the level the wave oscillates around.

The period is the time for one full cycle. The three parameters come from the maximum, minimum and period.

  • The amplitude is a=hmax−hmin2a=\frac{h_\mathrm{max}-h_\mathrm{min}}{2}a=2hmax​−hmin​​.
  • The midline value is d=hmax+hmin2d=\frac{h_\mathrm{max}+h_\mathrm{min}}{2}d=2hmax​+hmin​​.
  • For period TTT, the parameter is b=2πTb=\frac{2\pi}{T}b=T2π​.

Check the model against one known point from the data. If the prediction differs, recheck the period and the starting position.

Worked example

Building a sinusoidal model

The depth of water in a harbour varies between 2.0 m2.0\,\mathrm{m}2.0m and 6.0 m6.0\,\mathrm{m}6.0m with a period of 12 hours. At t=0t=0t=0 the depth is on the midline and rising. Find a model and the depth at t=1t=1t=1.
  1. 1
    a=6.0−2.02=2.0a=\frac{6.0-2.0}{2}=2.0a=26.0−2.0​=2.0 and d=6.0+2.02=4.0d=\frac{6.0+2.0}{2}=4.0d=26.0+2.0​=4.0.
  2. 2
    b=2π12=π6b=\frac{2\pi}{12}=\frac{\pi}{6}b=122π​=6π​.
  3. 3
    The model is h=2.0sin⁡(πt6)+4.0h=2.0\sin\left(\frac{\pi t}{6}\right)+4.0h=2.0sin(6πt​)+4.0.
  4. 4
    At t=1t=1t=1: h=2.0sin⁡(π6)+4.0=2.0×0.5+4.0=5.0h=2.0\sin\left(\frac{\pi}{6}\right)+4.0=2.0\times0.5+4.0=5.0h=2.0sin(6π​)+4.0=2.0×0.5+4.0=5.0.

Answer: h=2.0sin⁡(πt6)+4.0h=2.0\sin\left(\frac{\pi t}{6}\right)+4.0h=2.0sin(6πt​)+4.0, and the depth at t=1t=1t=1 is 5.0 m5.0\,\mathrm{m}5.0m.

Key terms · tap one for its definition

amplitude

The amplitude of a sinusoidal function is half the difference between its maximum and minimum values.

period

The period of a sinusoidal function is the length of one complete cycle, measured in the units of the input.

Part 3 done

4

Part 4 of 6: A logistic model levels off at a limit

Some models are used far into the future, so you need to know where they head. An exponential model grows without bound. A logistic model slows down and approaches a fixed value, which is a limit.

In P=L1+Ce−ktP=\frac{L}{1+Ce^{-kt}}P=1+Ce−ktL​, with positive CCC and kkk, the term e−kte^{-kt}e−kt shrinks towards 0 as ttt increases. The denominator then approaches 1. So PPP approaches LLL. The initial value is found by setting t=0t=0t=0.

Worked example

Reading a logistic model

A population is modelled by P=5001+9e−0.4tP=\frac{500}{1+9e^{-0.4t}}P=1+9e−0.4t500​, where ttt is in years. Find the initial population and the limit as ttt grows.
  1. 1
    At t=0t=0t=0, e0=1e^{0}=1e0=1, so P=5001+9=50P=\frac{500}{1+9}=50P=1+9500​=50.
  2. 2
    As ttt grows, e−0.4te^{-0.4t}e−0.4t approaches 0.
  3. 3
    The denominator approaches 1, so PPP approaches 500.

Answer: The initial population is 50, and the limit is 500.

The curve never reaches the limit at a finite time. It gets closer to 500 but stays below it.

Figure 1. Both curves start at 50 and grow at first. The logistic curve flattens towards its limit, while the exponential curve keeps rising.
Go deeperWhy the curve is S-shaped

A logistic model assumes growth slows as the quantity approaches LLL, because resources run short. Early on there is plenty of room, so growth looks exponential.

Growth is fastest when the quantity is half of LLL. After that point the curve bends over and flattens towards the limit.

Key term

limit

A limit is the value a function approaches as its input approaches a specified value.

Part 4 done

Quick check

One question on the part you just read

Easy

In the logistic model P=L1+Ce−ktP=\frac{L}{1+Ce^{-kt}}P=1+Ce−ktL​, what does LLL represent?
Show answer and explanation

Answer: C · Worked explanation

As ttt grows, e−kte^{-kt}e−kt approaches 0, so the denominator approaches 1 and PPP approaches LLL. The value LLL is therefore the limit of the model. The initial value is L1+C\frac{L}{1+C}1+CL​, not LLL, and the rate of growth is controlled by kkk.
5

Part 5 of 6: Comparing models means judging both fit and sense

Two models can describe the same data, so you need ways to compare them. Start with the residuals, which are observed values minus predicted values. Small residuals on both sides of zero suggest a good fit. Residuals of one sign mean the model is consistently too low or too high.

  • Fit: compare the size of the residuals on the same data.
  • Behaviour: check that predictions stay sensible outside the data.
  • Simplicity: prefer the simpler model when the fits are similar.
  • Assumptions: check that the context supports them.

No single criterion decides. A model with the smallest residuals may still give impossible values, such as a negative population.

Common mistake

A good fit does not prove that a model is the true rule. Different functions can fit the same data over a short range.

Part 5 done

6

Part 6 of 6: Every model has limitations that should be stated

After choosing a model, say where it stops working. A limitation is a condition under which the model’s predictions become unreliable or invalid. Most limitations fall into a few types.

  • Data range: predictions far outside the data may fail.
  • Simplified assumptions: a constant rate or amplitude may not hold.
  • Type of quantity: a continuous model may give 3.6 people.
  • Impossible values: lengths and populations cannot be negative.

When you state a limitation, name it precisely and link it to the situation. For example, a tide model ignores weather, so it cannot predict unusual water levels caused by storms.

Key term

limitation

A limitation is a condition or situation in which a model’s predictions become unreliable or invalid.

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[1 mark]

    State the type of function best suited to a quantity that grows quickly at first and then levels off at an upper value.

    Show mark schemeHide mark scheme

    Model answer

    A logistic function.

    • 1 mark for stating a logistic function
  2. 2.Determine[3 marks]

    Determine aaa, ddd and bbb in the model h=asin⁡(bt)+dh=a\sin(bt)+dh=asin(bt)+d for a depth that varies between 1.0 m1.0\,\mathrm{m}1.0m and 5.0 m5.0\,\mathrm{m}5.0m with a period of 8 hours.

    Show mark schemeHide mark scheme

    Model answer

    a=5.0−1.02=2.0a=\frac{5.0-1.0}{2}=2.0a=25.0−1.0​=2.0. d=5.0+1.02=3.0d=\frac{5.0+1.0}{2}=3.0d=25.0+1.0​=3.0. b=2π8=π4b=\frac{2\pi}{8}=\frac{\pi}{4}b=82π​=4π​.

    • 1 mark for obtaining a=2.0a=2.0a=2.0
    • 1 mark for obtaining d=3.0d=3.0d=3.0
    • 1 mark for obtaining b=π4b=\frac{\pi}{4}b=4π​
  3. 3.Determine[2 marks]

    Determine the initial population and the limit of PPP as ttt grows, given that P=8001+4e−0.3tP=\frac{800}{1+4e^{-0.3t}}P=1+4e−0.3t800​.

    Show mark schemeHide mark scheme

    Model answer

    At t=0t=0t=0, P=8001+4=160P=\frac{800}{1+4}=160P=1+4800​=160. As ttt grows, e−0.3te^{-0.3t}e−0.3t approaches 0, so PPP approaches 800.

    • 1 mark for obtaining an initial population of 160
    • 1 mark for stating that the limit is 800
  4. 4.Compare[3 marks]

    Compare the long-term behaviour of an exponential growth model and a logistic model.

    Show mark schemeHide mark scheme

    Model answer

    An exponential growth model keeps increasing without bound. A logistic model slows as it approaches a limit. So the logistic model suits quantities with a maximum, such as a population with limited food.

    • 1 mark for stating that exponential growth has no upper bound
    • 1 mark for stating that a logistic model approaches a limit
    • 1 mark for linking the logistic model to a quantity with a maximum
  5. 5.Discuss[3 marks]

    Discuss one limitation of using a sinusoidal model for the depth of water in a harbour over several months.

    Show mark schemeHide mark scheme

    Model answer

    The model assumes a constant amplitude and period. Real depths are affected by weather, which the model ignores. Predictions far from the data used to build it may therefore be unreliable.

    • 1 mark for stating that the model assumes constant amplitude or period
    • 1 mark for identifying an effect the model ignores, such as weather
    • 1 mark for concluding that predictions may be unreliable
  6. 6.Outline[3 marks]

    Outline how you would decide between two models for the same data.

    Show mark schemeHide mark scheme

    Model answer

    Compare the residuals of each model on the same data. Check that predictions make sense in context, with no impossible values. Prefer the simpler model if the fits are similar.

    • 1 mark for stating that residuals are compared
    • 1 mark for stating that predictions are checked for sense in context
    • 1 mark for stating that the simpler model is preferred when fits are similar

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 situation is best modelled by a piecewise function?
Show answer and explanation

Answer: A · Worked explanation

A price with a different rate above 10 kg uses one rule up to 10 kg and another rule beyond. That is a piecewise function. Tides suit a sinusoidal model, limited food suits a logistic model and cooling suits an exponential model.
4.

Medium·Determine[2 marks]

A temperature varies between 8 ∘C8\,^{\circ}\mathrm{C}8∘C and 20 ∘C20\,^{\circ}\mathrm{C}20∘C. In the model y=asin⁡(bx)+dy=a\sin(bx)+dy=asin(bx)+d, determine aaa and ddd.
Show answer and explanation

Answer: B · Worked explanation

The amplitude is half the difference: 20−82=6\frac{20-8}{2}=6220−8​=6. The midline value is the mean of the maximum and minimum: 20+82=14\frac{20+8}{2}=14220+8​=14. The value 12 is the full difference between maximum and minimum, so using it as the amplitude doubles the true amplitude. Using 12 for ddd confuses the difference with the mean.
5.

Medium·Determine[2 marks]

A model y=asin⁡(bt)+dy=a\sin(bt)+dy=asin(bt)+d has a period of 10 hours. Determine bbb.
Show answer and explanation

Answer: A · Worked explanation

The period and bbb are related by b=2πTb=\frac{2\pi}{T}b=T2π​, so b=2π10=π5b=\frac{2\pi}{10}=\frac{\pi}{5}b=102π​=5π​. The value 5π5\pi5π multiplies instead of dividing. The value 5π\frac{5}{\pi}π5​ is the reciprocal of the correct answer, which comes from using T=2πbT=\frac{2\pi}{b}T=b2π​ the wrong way round. A value of 10 simply repeats the period.
6.

Medium·Calculate[2 marks]

A population is modelled by P=6001+5e−0.5tP=\frac{600}{1+5e^{-0.5t}}P=1+5e−0.5t600​. Calculate the initial population.
Show answer and explanation

Answer: B · Worked explanation

At t=0t=0t=0, e0=1e^{0}=1e0=1, so P=6001+5×1=6006=100P=\frac{600}{1+5\times1}=\frac{600}{6}=100P=1+5×1600​=6600​=100. The value 120 comes from forgetting the 1 in the denominator and dividing by 5. The value 600 is the limit of the model as ttt grows, not the initial value, and the value 1 is only a term in the denominator.
7.

Medium·Determine[2 marks]

A drink cools according to T=60(0.9)t+20T=60(0.9)^t+20T=60(0.9)t+20, where TTT is in ∘C^{\circ}\mathrm{C}∘C and ttt is in minutes. Determine the value that TTT approaches as ttt grows.
Show answer and explanation

Answer: B · Worked explanation

As ttt grows, 0.9t0.9^t0.9t approaches 0, so TTT approaches 60×0+20=2060\times0+20=2060×0+20=20, which is 20 ∘C20\,^{\circ}\mathrm{C}20∘C. This is the temperature of the surroundings. The value 80 is the starting temperature at t=0t=0t=0. The value 0 forgets the constant term 20, and 60 is only the size of the initial gap.
8.

Hard[2 marks]

An exponential model P=50(1.3)tP=50(1.3)^tP=50(1.3)t fits a colony’s data well for t=0t=0t=0 to 6. A student uses it to predict the colony in a small dish at t=40t=40t=40. Why is this unreliable?
Show answer and explanation

Answer: B · Worked explanation

The prediction is far outside the data and the model assumes growth without bound. A dish has finite food and space, so growth would level off, as in a logistic model. Exponential functions can be fitted to data, and an initial value of 50 is not a flaw.
9.

Hard[2 marks]

Model X has a largest residual of 0.4 but predicts negative populations in part of the domain being studied. Model Y has a largest residual of 0.6 and positive values throughout. Which statement is best supported?
Show answer and explanation

Answer: D · Worked explanation

A model must fit the data and also make sense in context. A negative population is impossible, so X has a serious limitation despite its smaller residuals. Residuals are not the only criterion. No model needs zero residuals to be useful, so rejecting both is wrong.
10.

Hard·Determine[2 marks]

A depth is modelled by h=2sin⁡(πt6)+4h=2\sin\left(\frac{\pi t}{6}\right)+4h=2sin(6πt​)+4, where ttt is in hours. The depth is 5 at t=1t=1t=1. Determine the next time, after t=1t=1t=1, at which the depth is 5.
Show answer and explanation

Answer: C · Worked explanation

The period is 12 hours and the maximum occurs at t=3t=3t=3. The curve is symmetric about the maximum, so the next time is 3+(3−1)=53+(3-1)=53+(3−1)=5. The value 13 adds a full period, which gives a later time, not the next one. The value 11 subtracts 1 from the period.

Keep going

Next topic2.10 Scaling and linearizing dataPrevious topic2.8 Transformations of graphs

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

0/6 read

  1. 1Different situations call for different families of function
  2. 2The pattern in the data suggests a suitable model
  3. 3Key features of the data give the parameters
  4. 4A logistic model levels off at a limit
  5. 5Comparing models means judging both fit and sense
  6. 6Every model has limitations that should be stated
  7. Check your recall
  8. 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

Logistic and exponential growth

Population051015200100200300400500Time (years)
  • Logistic model
  • Limit, 500
  • Exponential, from 50