MATH 1 • MODELING & APPLICATIONS

Choosing Linear vs. Exponential — I can compare linear vs exponential models for the same situation and justify the choice.

Learn to identify growth patterns in real data and select the model that truly fits.

Historical Context & Motivation

For centuries, mathematicians and scientists have needed to describe how quantities change over time. Some quantities grow at a steady pace — a merchant earning a fixed profit each month, for instance — while others snowball, doubling again and again until the numbers become staggering. The distinction between these two patterns, linear growth and exponential growth, is one of the most important ideas in mathematical modeling. Understanding which type of change is at work determines whether your predictions will be trustworthy or wildly off.

~300 BCE
Euclid & Arithmetic Sequences
Euclid studied arithmetic sequences — lists of numbers that increase by a constant amount — laying the groundwork for what we now call linear relationships.
1202
Fibonacci & Compound Growth
Leonardo of Pisa (Fibonacci) explored problems involving breeding rabbits and compound interest, revealing patterns of multiplicative growth that underlie exponential models.
1798
Malthus on Population
Thomas Malthus argued that population grows exponentially while food supply grows linearly — a clash of two models that shaped economics and policy debates for centuries.
2020
COVID-19 & Model Choice
During the early pandemic, health officials needed to decide whether case counts were rising linearly or exponentially. Choosing the wrong model led to dangerously inaccurate forecasts, proving that model selection has life-or-death stakes.

The central question is deceptively simple: given a table of data or a real-world scenario, how do you decide whether a linear model or an exponential model is the better fit — and how do you justify that choice with mathematical evidence? This lesson will give you the tools to answer confidently.

Core Principles & Definitions

Before you can choose between the two models, you need to understand what makes each one tick. Both models describe how an output quantity changes as an input quantity (usually time) increases, but they do so in fundamentally different ways.

1

Constant Rate of Change

In a linear model, the output changes by the same fixed amount for every equal step in the input. If you add 3 each time, you always add 3 — no matter how large the output gets.
2

Constant Ratio of Change

In an exponential model, the output is multiplied by a fixed factor for every equal step. Doubling every hour means you multiply by 2, whether the output is 10 or 10,000.
3

First Differences vs. Ratios

Check for linearity by computing first differences (subtract consecutive outputs). Check for exponential behavior by computing consecutive ratios (divide consecutive outputs). If one of these is constant, you've found your model.
4

The Graph Test

A linear function produces a straight line. An exponential function produces a curve that starts slowly and then accelerates (growth) or drops quickly and levels off (decay).
KEY TAKEAWAY
Think of a linear model like climbing a staircase — you go up by the same height with every step. An exponential model is more like a rumor spreading through school: each person who hears it tells two more, so the count doesn't just grow — it multiplies. To figure out which pattern fits your data, ask: "Am I adding the same amount each time, or multiplying by the same factor?"

Visual Explanation — Linear vs. Exponential Side by Side

The most powerful way to see the difference between these two models is to plot them on the same set of axes. The diagram below shows a linear function and an exponential function that start at the same point. Notice how the linear graph rises at a steady pace while the exponential curve appears to "take off" dramatically.

Both functions start at y = 10 when x = 0. The blue linear line rises by a constant 24 each step, while the violet exponential curve doubles each step. By x = 5, the exponential value has far outpaced the linear one.

At small x-values, the two graphs look surprisingly similar — this is why it can be tricky to distinguish them from limited data. But as x increases, the exponential curve pulls away dramatically. If you only look at the first few data points, you might mistake exponential growth for linear growth. That's exactly why you need the first-differences and ratios tests described in the next sections.

Mathematical Framework

Each model type has a standard equation form. Recognizing these forms — and knowing what each variable represents — is the first step toward writing and comparing equations for real situations.

LINEAR MODEL
y = mx + b
m = slope (constant rate of change, or how much y changes for each unit increase in x); b = y-intercept (the starting value when x = 0).
EXPONENTIAL MODEL
y = a · bˣ
a = initial value (the output when x = 0); b = common ratio (the factor by which y is multiplied for each unit increase in x). When b > 1, the model represents growth; when 0 < b < 1, it represents decay.

The Diagnostic Tests

FIRST DIFFERENCES (LINEAR TEST)
Δy = y₂ − y₁
Compute the difference between consecutive outputs. If Δy is constant for equally spaced inputs, the data is linear.
CONSECUTIVE RATIOS (EXPONENTIAL TEST)
r = y₂ ÷ y₁
Compute the ratio of each output to the previous one. If r is constant for equally spaced inputs, the data is exponential.
⚠️ Important Note
These tests require equally spaced inputs. If the x-values jump by 1, then by 3, then by 2, you must adjust your differences or ratios accordingly — or interpolate first. For most classroom problems, the x-values will be evenly spaced.

Analyzing Data — First Differences & Ratios in Action

Let's see the diagnostic tests applied to two different data sets. The first table represents the number of pages a student reads each day; the second represents bacteria in a petri dish counted every hour.

Data Set A: Pages read per day
Day (x)Pages Read (y)First Difference (Δy)Ratio (y₂ ÷ y₁)
020
135+151.75
250+151.43
365+151.30
480+151.23

In Data Set A, the first differences are constant at +15, while the ratios vary. This means the data is linear. The equation is y = 15x + 20.

Data Set B: Bacteria count per hour
Hour (x)Bacteria (y)First Difference (Δy)Ratio (y₂ ÷ y₁)
0100
1300+2003
2900+6003
32,700+1,8003
48,100+5,4003

In Data Set B, the first differences are growing rapidly, but the consecutive ratios are constant at 3. This means the data is exponential. The equation is y = 100 · 3ˣ.

This flowchart summarizes the decision process: compute both first differences and consecutive ratios, then check which column produces a constant value. If neither is constant, a different model type may be needed.

Worked Example — Saving Money Two Ways

Suppose you start with $200 in a savings account. Under Plan A, you deposit $50 every month. Under Plan B, the account earns 8% interest each month (compounded monthly, no additional deposits). Which plan is linear and which is exponential? Write equations for both and predict the balance after 12 months.

Two Savings Plans: Linear or Exponential?
1
Step 1 — Identify the Type of ChangePlan A adds a fixed $50 each month — that's a constant amount added, so Plan A is linear. Plan B multiplies the balance by 1.08 each month — that's a constant ratio, so Plan B is exponential.
2
Step 2 — Write the Linear Equation (Plan A)Using y = mx + b, the slope m = 50 (dollars per month) and the y-intercept b = 200 (the starting balance).
y = 50x + 200
3
Step 3 — Write the Exponential Equation (Plan B)Using y = a · bˣ, the initial value a = 200 and the common ratio b = 1.08 (since 8% growth means multiplying by 1 + 0.08).
y = 200 · 1.08ˣ
4
Step 4 — Predict at x = 12 MonthsPlan A: y = 50(12) + 200 = 600 + 200 = $800. Plan B: y = 200 · 1.08¹² = 200 · 2.518... ≈ $503.63.
After 12 months, Plan A gives $800 while Plan B gives ≈ $503.63.
5
Step 5 — Interpret and JustifyAt 12 months, the linear plan wins because the constant deposits add up quickly. However, if we extend the timeline, the exponential plan will eventually surpass the linear one — that's the nature of multiplicative growth. The choice of model isn't about which is 'better' in the abstract; it's about which accurately describes the situation's mechanics. Adding a fixed deposit is linear. Earning a percentage is exponential.

Strengths & Limitations of Each Model

No model is perfect for every situation. The table below compares linear and exponential models across several important criteria to help you decide which is appropriate — and to understand what each model cannot do.

CriterionLinear ModelExponential Model
Pattern of changeAdds or subtracts the same amount every periodMultiplies by the same factor every period
Graph shapeStraight lineJ-curve (growth) or ski-slope (decay)
Good for short-term predictionYes — reliable when the trend is steadyYes — if the ratio stays consistent
Long-term behaviorCan go negative, which may not be realisticGrows without bound or approaches zero; never crosses the x-axis
Common real-world usesFixed hourly wage, constant speed, even depreciationPopulation growth, radioactive decay, compound interest
Key limitationCannot capture accelerating or decelerating changeMay overestimate when real-world constraints slow growth
KEY TAKEAWAY
Choosing between linear and exponential models is like choosing between driving at a constant speed and pressing the gas pedal harder every second so that your speed keeps multiplying. Both describe motion, but they describe very different driving experiences. Always match the model to the actual mechanism behind the situation.

Connections to Advanced Modeling

Linear and exponential are the two simplest function families, but real-world data often requires more sophisticated models. In later math courses, you will encounter additional function types — each designed to capture patterns that linear and exponential models cannot.

FeatureLinear / Exponential (Math 1)Advanced Models (Math 2+)
Function typesy = mx + b and y = a · bˣQuadratic, logarithmic, logistic, polynomial, piecewise
Number of parametersTwo (slope & intercept, or initial value & ratio)Three or more, allowing the curve to bend, peak, or level off
Fitting methodFirst differences, ratios, or two-point calculationRegression analysis (least-squares, technology-assisted)
When neededSteady growth or constant-ratio changeData that accelerates and then slows, peaks, or has limits

One model worth knowing about now is the logistic model, which starts out looking exponential but then levels off at a carrying capacity. Think of a social media post going viral: at first shares multiply rapidly (exponential), but eventually everyone has already seen it and growth slows. For Math 1, though, you'll focus on recognizing and justifying the choice between linear and exponential.

Practice Problems

PROBLEM 1CONCEPTUAL
A friend says, "Both linear and exponential functions can start at the same point and increase, so it doesn't matter which one I use." Explain why this reasoning is flawed, using the idea of rate of change in your answer.
PROBLEM 2BASIC CALCULATION
Given the data table below, determine whether the relationship is linear or exponential. Then write the equation. x: 0, 1, 2, 3, 4 y: 5, 15, 45, 135, 405
PROBLEM 3INTERMEDIATE
A car's value decreases over time. The table shows: Year: 0, 1, 2, 3 Value ($): 20,000, 17,000, 14,450, 12,282.50 Is the depreciation linear or exponential? Write the equation and predict the value after 5 years.
PROBLEM 4APPLIED
A lake's fish population is tracked for several years. A biologist proposes two models: Model L gives P = 200 + 80t and Model E gives P = 200 · 1.3ᵗ, where t is years since observation began. At t = 3, the actual count is 439 fish. Which model fits better at t = 3? Justify your answer and explain which model you would trust for long-term prediction.
PROBLEM 5CRITICAL THINKING
A student collects data and finds that first differences are approximately constant at +12 (± 1) and consecutive ratios are approximately constant at 1.04 (± 0.005). The student is confused because both tests seem to "pass." Explain why this can happen, describe what additional analysis the student should perform, and state which model you would recommend.

Summary — Choosing Linear vs. Exponential

A linear model (y = mx + b) describes situations where the output changes by a constant amount for each equal step in the input, producing a straight-line graph. An exponential model (y = a · bˣ) describes situations where the output is multiplied by a constant factor each step, producing a curve that either grows rapidly or decays toward zero. The key diagnostic tools are first differences (constant for linear data) and consecutive ratios (constant for exponential data).

To justify your model choice, show numerical evidence from the data (constant differences or constant ratios) and connect it to the real-world mechanism — is the quantity growing by addition or by multiplication? When in doubt, extend your data range, plot the points, and look for curvature. Mastering this decision prepares you for more advanced modeling with quadratic, logistic, and other function families in future courses.

Varsity Tutors • Math 1 • Choosing Linear vs. Exponential