Analyzing Functions and Models

Analyzing Functions and Models asks you to move beyond simply graphing or evaluating a function and instead interrogate what its behavior reveals about a real situation. This work extends the broader toolkit developed in Advanced Analysis Methods, where identifying structure and drawing evidence-based conclusions are central skills. Here, the "text" under analysis is a function, and the "argument" it makes is a claim about how one quantity changes in relation to another.

Before a function can be used as a model, you need to be able to describe how it behaves. Analysis typically begins with four questions: What is the domain and range? Where does the graph cross the axes? Over which intervals is the function increasing, decreasing, or constant? And what happens to the output as the input grows very large or very small (the end behavior)?

The average rate of change over an interval \( [a, b] \) is calculated as \( \dfrac{f(b) - f(a)}{b - a} \). For a linear function this value is constant everywhere, which is exactly what makes a straight line an appropriate model for a process changing at a fixed rate. For nonlinear functions, the rate of change itself varies, and comparing it across intervals tells you whether a process is accelerating, decelerating, or leveling off toward an asymptote.

A model is only useful if its structure matches the pattern in the data or context. A situation with a constant rate of change calls for a linear model \( f(x) = mx + b \). A situation with a constant percent change per unit of input, such as compounding or decay, calls for an exponential model \( f(x) = a(1+r)^x \). A situation involving a single maximum or minimum, such as height under gravity, calls for a quadratic model \( f(x) = ax^2 + bx + c \). Logarithmic models arise when growth slows sharply as the input increases, since a logarithm reverses the effect of exponential change.

Choosing among these forms is itself an analytical judgment: you examine successive differences or ratios in a data set, or listen for context clues such as "grows by three percent each year" versus "decreases by a fixed amount each hour," and select the family of functions whose defining behavior matches.

Once a model is built, each parameter should translate into a real statement about the situation. In \( f(x) = mx + b \), the slope \( m \) is the rate at which the output changes per unit of input, and \( b \) is the starting value when \( x = 0 \). In an exponential model \( f(x) = a(1+r)^x \), \( a \) is the initial amount and \( r \) is the constant growth or decay rate. In a quadratic model, the vertex identifies the maximum or minimum value the process reaches, found at \( x = -\dfrac{b}{2a} \).

Context also restricts the domain of a model. A function describing the height of a thrown object is only meaningful while the object is airborne, and a function describing a population is only meaningful for non-negative time and non-negative output. Ignoring these restrictions leads to conclusions the model was never built to support.

Suppose a ball is launched from a platform and its height in feet after \( t \) seconds is modeled by \( h(t) = -16t^2 + 64t + 5 \). Analyzing this function means asking what its structure tells you about the flight.

The negative leading coefficient means the graph opens downward, so the function increases and then decreases, matching the physical rise and fall of the ball. The vertex occurs at \( t = -\dfrac{64}{2(-16)} = 2 \), so the ball reaches its maximum height two seconds after launch. Evaluating \( h(2) = -16(2)^2 + 64(2) + 5 = 69 \) shows that maximum height is 69 feet. The value \( h(0) = 5 \) confirms the ball started 5 feet above the ground, consistent with a launch from a platform rather than the ground itself. Because height cannot be negative in this context, the practical domain of the model ends the moment \( h(t) \) returns to zero, not at every value where the algebraic expression is defined.

Domain: The complete set of input values for which a function or model is defined and meaningful.

Range: The complete set of output values a function can produce over its domain.

End Behavior: The trend a function's output follows as the input increases or decreases without bound.

Rate of Change: The ratio describing how much the output of a function changes relative to a change in input.

Vertex: The maximum or minimum point on the graph of a quadratic function.

Exponential Model: A function of the form \( f(x) = a(1+r)^x \) used to represent constant percent growth or decay.

Asymptote: A line that a function's graph approaches but never actually reaches as the input grows extreme.

Beyond projectile motion, the same analytical process applies to population growth, cooling temperatures, investment accounts, and depreciation of an asset. In each case, you identify the pattern of change, select a matching function family, fit its parameters to the given information, and then interpret the result within the constraints of the real situation, rather than treating the equation as an abstract object detached from what it represents.

These modeling decisions connect closely to the reasoning developed in Analytical Problem Solving Strategies, where breaking a complex problem into identifiable structures is the first step toward a workable solution.

Related lessons

• Advanced Analysis Methods