Selecting Function Families — I can select an appropriate...

SECTION 1

Historical Context & Motivation

Mathematics has always been a language for describing the world around us. From ancient astronomers tracking planetary orbits to modern scientists modeling population growth, the central challenge has remained the same: which type of equation best captures the pattern we observe? Over the centuries, mathematicians developed distinct families of functions, each tailor-made for certain kinds of behavior. Understanding this history helps us see why we have so many function types and when each one shines.

~300 BCE

Polynomials in Geometry

Greek mathematicians like Euclid used polynomial relationships — areas of squares (x²), volumes of cubes (x³) — to solve geometric problems, laying the groundwork for polynomial functions.

1614

Napier Invents Logarithms

John Napier published tables of logarithms to simplify astronomical calculations. This introduced the logarithmic function family as the inverse of exponential growth.

1798

Malthus & Exponential Growth

Thomas Malthus argued that populations grow exponentially while food supply grows linearly. His work popularized the use of exponential functions in social science and biology.

1822

Fourier & Trigonometric Modeling

Joseph Fourier showed that any periodic phenomenon — heat waves, sound, tides — can be modeled using combinations of trigonometric functions (sines and cosines).

20th Century

Modern Mathematical Modeling

With computers, scientists began fitting rational functions and other complex models to data. Selecting the right function family became a core skill in applied mathematics, engineering, and data science.

The big question that ties all of this history together is deceptively simple: given a real-world situation, how do you decide which function family to use? That's exactly the skill this lesson builds. By the end, you'll be able to look at a context — bacterial growth, ocean tides, the path of a basketball — and select the right model with a clear justification.

SECTION 2

Core Principles & Function Family Profiles

Before you can choose the right model, you need to know the main characters. Each function family has a distinctive shape, a signature behavior, and a set of real-world contexts where it naturally appears. Think of each family as a tool in a toolbox — a hammer is great for nails, but you wouldn't use it to turn a screw. The same logic applies to mathematical models.

Polynomial Functions

Built from terms like x, x², x³. They model projectile paths (quadratic), volume relationships (cubic), and smooth curves with a finite number of turns. Key clue: the data eventually rises or falls without bound in the same direction at both ends (even degree) or opposite directions (odd degree).

Rational Functions

Ratios of two polynomials (e.g., 1/x). They feature asymptotes — values the output approaches but never reaches. Great for modeling rates, concentrations that level off, or situations where a quantity is inversely related to another.

Exponential Functions

Form: a · bˣ. Output multiplies by a constant factor each step. Constant percentage change is the hallmark. Models compound interest, radioactive decay, viral spread, and population growth.

Logarithmic Functions

The inverse of exponentials: y = log(x). Growth that is rapid at first but slows dramatically. Appears in the Richter scale (earthquakes), decibels (sound), and the learning curve phenomenon.

Trigonometric Functions

Sine, cosine, and tangent produce periodic (repeating) patterns. If the data cycles — temperatures across seasons, hours of daylight, tides, heartbeats — a trig function is your best bet.

✦ KEY TAKEAWAY

Choosing a function family is like choosing a vehicle for a trip. Need to cross an ocean? A car won't help — you need a boat (or a plane). Similarly, if your data repeats in cycles, an exponential function won't capture that; you need a trigonometric model. The context tells you what 'vehicle' to pick, and the behavior of the data confirms you chose correctly.

SECTION 3

Visual Comparison of Function Families

Seeing all five function families side by side makes their differences unmistakable. The diagram below plots a representative curve from each family on the same coordinate grid. Pay attention to end behavior (what happens as x gets very large or very negative), symmetry, and rate of change — these are the visual fingerprints that help you identify the right family.

Notice how the exponential curve shoots upward much faster than the polynomial, while the logarithmic curve flattens out. The trigonometric curve is the only one that repeats its pattern.

Each curve has a unique visual fingerprint. The polynomial (blue) forms a smooth U-shape that is symmetric about its vertex. The exponential (pink) starts nearly flat but accelerates sharply upward — a sign of constant percentage growth. The logarithmic (amber) is the mirror image: rapid initial rise that tapers off. The rational function (violet, dashed) drops toward a horizontal line it never quite reaches, revealing an asymptote. Finally, the trigonometric sine wave (cyan) oscillates above and below the axis in a perfectly repeating cycle. When you encounter data in the real world, ask yourself: which of these shapes does my data most resemble?

SECTION 4

Mathematical Framework — General Forms & Key Features

Each function family has a general algebraic form. Knowing these forms helps you match context clues to the right equation. Below are the standard forms along with the features that distinguish each family.

POLYNOMIAL (DEGREE n)

f(x) = aₙxⁿ + aₙ₋₁xⁿ⁻¹ + … + a₁x + a₀

Here n is the degree (highest exponent). A quadratic has n = 2, a cubic has n = 3. The degree controls how many times the curve can change direction (at most n − 1 turns).

EXPONENTIAL

f(x) = a · bˣ (b > 0, b ≠ 1)

a is the initial value (y-intercept when x = 0). b is the growth factor: b > 1 means growth, 0 < b < 1 means decay. The output multiplies by b every time x increases by 1.

LOGARITHMIC

f(x) = a · log_b(x) + c (x > 0)

The logarithm is the inverse of the exponential. It answers the question: 'To what power must b be raised to get x?' Output grows without bound but does so more and more slowly.

TRIGONOMETRIC (SINUSOIDAL)

f(x) = A · sin(Bx + C) + D

A = amplitude (height of the wave), B controls the period (how quickly it repeats: period = 2π / |B|), C shifts the wave left or right, and D shifts it up or down.

📐 Rational Functions — Quick Note

A rational function has the form f(x) = P(x) / Q(x), where P and Q are both polynomials. Wherever Q(x) = 0, the function is undefined, which creates vertical asymptotes. The behavior of the function as x → ±∞ is controlled by comparing the degrees of P and Q, which can create horizontal asymptotes.

SECTION 5

Decision Flowchart — Matching Context to Function

When you face a modeling problem, you can follow a systematic decision process. Ask yourself a series of diagnostic questions about the data or the situation. The flowchart below walks you through the key questions and points you toward the right function family.

Start at the top and work your way down. Each 'YES' sends you to a function family; each 'NO' moves you to the next question. This flowchart covers the most common scenarios you'll see in Grade 11.

This flowchart is a starting point, not a rigid rulebook. Some contexts could involve combinations — for instance, a population that grows exponentially at first but levels off might call for a logistic function (a topic for a later course). For Grade 11, focusing on the five core families will handle the vast majority of modeling scenarios.

Common real-world contexts and their matching function families

Context Clue	Function Family	Why?
Data doubles every fixed interval	Exponential	Constant multiplicative rate → y = a · bˣ
Ball thrown upward; max height then falls	Polynomial (quadratic)	Parabolic path from constant acceleration due to gravity
Temperature oscillates daily/yearly	Trigonometric	Regular, predictable repeating cycle
Speed of learning: fast then plateaus	Logarithmic	Rapid early growth that slows over time
Average cost per unit as production increases	Rational	Cost per unit → asymptote as fixed costs are spread thinner

SECTION 6

Worked Example — Selecting and Justifying a Function Family

Let's work through a full modeling problem from start to finish. The goal is not just to pick a function family, but to justify your choice with clear reasoning tied to the context and the data's behavior.

🧪 Problem

A biologist observes a colony of bacteria. She records the population every hour and gets the following data: Hour 0 → 500 cells, Hour 1 → 1 000, Hour 2 → 2 000, Hour 3 → 4 000, Hour 4 → 8 000. She wants to write a function to predict the population at hour t. Which function family should she use? Write a possible equation and justify.

Solution: Bacterial Growth Model

Step 1 — Identify the Pattern in the DataLook at what happens to the population each hour. From 500 to 1 000 is a factor of 2. From 1 000 to 2 000 is again a factor of 2. From 2 000 to 4 000 — factor of 2. The population doubles every hour. This is a constant multiplicative change, not a constant additive change.

Growth factor = 2 per hour (constant ratio)

Step 2 — Match the Pattern to a Function FamilyA constant multiplicative change is the defining feature of an exponential function. If the data grew by a constant amount each hour (e.g., +500 each time), we would use a linear (polynomial degree 1) model instead. But multiplying by 2 each step means exponential.

Function family: Exponential

Step 3 — Write the EquationThe general exponential form is f(t) = a · bᵗ. Here, a = 500 (initial population at t = 0) and b = 2 (the population doubles). So the model is:

P(t) = 500 · 2ᵗ

Step 4 — Verify with DataCheck: P(0) = 500 · 2⁰ = 500 ✓. P(3) = 500 · 2³ = 500 × 8 = 4 000 ✓. P(4) = 500 · 2⁴ = 500 × 16 = 8 000 ✓. The model matches every data point.

All five data points verified ✓

Step 5 — Write the JustificationA full justification might sound like this: 'I selected an exponential model because the bacteria population doubles each hour, which is a constant multiplicative change. The data shows a ratio of 2 between consecutive outputs, not a constant difference. This matches the form f(t) = a · bᵗ, where a = 500 and b = 2. A polynomial model would not capture the accelerating growth, and a trigonometric model would not apply because the population does not cycle.'

Justification connects the constant ratio in the data to the exponential function's defining property.

SECTION 7

Strengths & Limitations of Each Family

No function family is perfect for every situation. Understanding the strengths and limitations of each type will help you avoid common modeling pitfalls. The table below summarizes what each family does well and where it breaks down.

Comparison of strengths and limitations across function families

Family	Strengths	Limitations
Polynomial	Very flexible; can fit many shapes. Easy to manipulate algebraically. Models projectiles, area, and revenue well.	End behavior always goes to ±∞, so long-term predictions can be unrealistic. Can overfit with high degrees.
Rational	Captures asymptotic behavior naturally. Great for rates, concentrations, and inverse relationships.	Undefined at certain x-values (vertical asymptotes). Harder to solve algebraically. Domain restrictions require attention.
Exponential	Perfect for constant-percentage growth/decay. Models compound interest, radioactive decay, and population dynamics.	Predicts unlimited growth, which is unrealistic long-term for populations. Doesn't model cycles or leveling off.
Logarithmic	Models diminishing returns well. Useful for measuring scales (decibels, Richter). Inverse of exponential.	Only defined for positive inputs (x > 0). Growth slows so much it may underpredict in some contexts.
Trigonometric	Only family that naturally repeats. Ideal for anything periodic — seasons, tides, sound waves, rotation.	Assumes the pattern repeats exactly and indefinitely. Doesn't model growth or decay unless combined with another function.

✦ KEY TAKEAWAY

Think of each function family as a specialist, not a generalist. An exponential function is like a sprinter — phenomenal at capturing explosive growth, but it can't run a marathon (it doesn't level off or repeat). A trigonometric function is like a metronome — perfectly rhythmic but unable to speed up or slow down. The best model is the one whose built-in behavior matches the real-world behavior you're describing.

SECTION 8

Connection to Advanced Modeling

The five function families you've learned here form the foundation for more sophisticated models you'll encounter in future courses. In AP Calculus, you'll study how these functions change using derivatives. In statistics, you'll use regression analysis to find the best-fitting function from data. And in advanced modeling, you'll combine families to create hybrid models like the logistic function (exponential growth that levels off) or damped oscillation (a trig wave multiplied by a decaying exponential).

How today's skills connect to future courses

What You Learn Now	Where It Leads
Identify exponential growth from constant ratios	Exponential regression in AP Statistics; differential equations in college math
Recognize periodic behavior → trig model	Fourier analysis in physics and engineering (breaking complex waves into sine components)
Use polynomial models for smooth, finite-turn curves	Taylor/Maclaurin series (using polynomials to approximate any function) in calculus
Identify asymptotic behavior → rational model	Limits and continuity in AP Calculus; pharmacokinetics (drug concentration over time)
Justify your function choice with reasoning	Model validation and residual analysis in data science

The skill of justifying your model choice is especially important in advanced work. In college and career contexts, people don't just want the answer — they want to know why you chose that model and what assumptions you're making. The reasoning habits you build here will serve you well whether you become an engineer, a biologist, a financial analyst, or a data scientist.

SECTION 9

Practice Problems

PROBLEM 1 — CONCEPTUAL

A student says, 'Since the data keeps going up, it must be exponential.' Explain why this reasoning is flawed. What additional information would you need to confirm that an exponential model is appropriate?

PROBLEM 2 — BASIC CALCULATION

A radioactive substance has a half-life of 3 years. Starting with 200 grams, which function family models the amount remaining after t years? Write the equation.

PROBLEM 3 — INTERMEDIATE

The average number of daylight hours in a city follows this approximate pattern: Jan → 9.5 h, Apr → 13.0 h, Jul → 14.5 h, Oct → 11.0 h, then back to ≈ 9.5 h in January. Which function family best models daylight hours as a function of the month? Explain two specific features of the data that support your answer.

PROBLEM 4 — APPLIED

A small company manufactures custom phone cases. They have $10 000 in fixed costs (rent, equipment) and each case costs $3 to produce. The average cost per case as a function of the number of cases produced, n, is C(n) = (10 000 + 3n) / n. Identify the function family, describe the long-term behavior of C(n), and explain what it means in context.

PROBLEM 5 — CRITICAL THINKING

A student collects data on the number of registered users on a new social media app: Month 1 → 100, Month 2 → 300, Month 3 → 850, Month 4 → 2 400, Month 5 → 6 500. She fits an exponential model and predicts 50 million users by Month 12. A classmate argues that a logistic model would be more realistic. Who is right, and why? What real-world factors should influence the choice?

SUMMARY

Lesson Summary

In this lesson, you learned to select from five major function families based on the behavior of a real-world context. Polynomial functions model smooth curves with a finite number of turns (projectiles, area, revenue). Rational functions feature asymptotes and model rates or concentrations that approach a limiting value. Exponential functions capture constant-percentage growth or decay — look for constant ratios between consecutive outputs. Logarithmic functions describe quantities that grow rapidly at first and then slow down dramatically. Trigonometric functions are the go-to choice whenever data repeats in a predictable cycle.

The most important skill is not just picking a family, but justifying your choice by connecting specific features of the data or context — constant ratios, periodicity, asymptotic behavior, number of turns — to the defining properties of the function family. Use the decision flowchart as a guide: ask whether the data cycles, grows by a constant percentage, levels off, approaches an asymptote, or follows a smooth polynomial curve. Always verify your model against the data and consider its limitations — every model is an approximation, and knowing where it breaks down is just as important as knowing where it works.