Using the Second Derivative Test to Determine Extrema - AP Calculus BC

SECTION 1

Historical Context & Motivation

The quest to find maximum and minimum values of functions is one of the oldest problems in mathematics, predating even the formal invention of calculus. Ancient Greek mathematicians such as Euclid and Apollonius sought to optimize geometric quantities—finding the shortest distance from a point to a line, or the largest rectangle inscribable in a given shape—using purely synthetic methods. These early optimization problems hinted at a deeper algebraic machinery that would not be fully realized for nearly two millennia. The eventual development of the Second Derivative Test represents the culmination of centuries of refinement in understanding how the curvature of a function's graph determines the nature of its turning points.

~300 BCE • Greek Optimization

Euclid and Apollonius solved extremal problems in geometry, such as finding the shortest path and the largest area inscribed in a figure, using purely geometric reasoning.

1665–1676 • Birth of Calculus

Newton and Leibniz independently developed the foundations of differential calculus, introducing the concept of the derivative and the idea that extrema occur where the tangent line is horizontal.

1740s • Euler's Systematization

Leonhard Euler formalized the use of higher-order derivatives to classify critical points, establishing the logical framework behind the second derivative test in his treatise Introductio in analysin infinitorum.

1797 • Lagrange's Théorie des fonctions analytiques

Joseph-Louis Lagrange provided rigorous Taylor-series-based arguments showing that the sign of the second derivative at a critical point determines whether the function curves upward or downward, completing the modern proof.

Modern Era • Standard Calculus Curriculum

The Second Derivative Test became a cornerstone of introductory calculus courses worldwide, valued for its elegance and direct connection between concavity and the nature of extrema.

The central question that the Second Derivative Test addresses is elegantly simple: once you have located a critical point where f′(c) = 0, how can you determine—without examining the sign of f′ on intervals to the left and right—whether that point is a local maximum, a local minimum, or neither? The answer lies in the concept of concavity, which the second derivative encodes directly. This test provides a computationally efficient alternative to the First Derivative Test, particularly when evaluating f″ at the critical point is simpler than analyzing the sign behavior of f′ across an interval.

SECTION 2

Core Principles & Definitions

Before applying the Second Derivative Test, you must have a firm command of the prerequisite concepts that underpin it. The test synthesizes three interrelated ideas: critical points, the behavior of the first derivative, and concavity as measured by the second derivative. Each of these ideas connects to the geometric shape of the curve in a way that makes classification of extrema nearly visual once the algebra is done.

Critical Point

A value x = c in the domain of f where f′(c) = 0 or f′(c) does not exist. Critical points are the only candidates for local extrema by Fermat's Theorem.

Concavity

A function is concave up on an interval where f″(x) > 0 (the curve bends upward like a cup) and concave down where f″(x) < 0 (the curve bends downward like a cap).

Second Derivative Test Statement

If f′(c) = 0 and f″(c) > 0, then f has a local minimum at c. If f′(c) = 0 and f″(c) < 0, then f has a local maximum at c. If f″(c) = 0, the test is inconclusive.

Inconclusive Case

When f″(c) = 0, the critical point could be a local max, local min, or an inflection point. You must revert to the First Derivative Test or use higher-order derivative tests.

✦

KEY TAKEAWAY

Think of a ball rolling along the curve of f. At a critical point, the ball momentarily has zero horizontal velocity (f′ = 0). If the ground curves upward (f″ > 0), the ball sits in a valley—a local minimum. If the ground curves downward (f″ < 0), the ball sits on a hilltop—a local maximum. If the ground is perfectly flat in both curvature and slope (f″ = 0), you cannot tell which way the ball will roll without further investigation.

SECTION 3

Visual Explanation

The following diagram illustrates the geometric intuition behind the Second Derivative Test. At each labeled critical point, the first derivative is zero—the tangent line is horizontal. The curvature of the graph at that point, captured by the sign of the second derivative, determines whether the point sits in a concave-up "bowl" (local minimum) or a concave-down "dome" (local maximum).

The purple curve shows a function with two critical points. At c₁ (pink), the curve is concave down (f″ < 0), producing a local maximum. At c₂ (cyan), the curve is concave up (f″ > 0), producing a local minimum. The dashed horizontal lines represent tangent lines with slope zero.

Observe how the concavity at each critical point determines the shape of the curve in a small neighborhood. Near the local maximum at c₁, the graph lies below its horizontal tangent line on both sides—this is precisely the geometric meaning of concave down. Near the local minimum at c₂, the graph lies above its horizontal tangent line on both sides—the hallmark of concave up. This relationship between concavity and the position of the curve relative to its tangent line is what makes the Second Derivative Test work.

SECTION 4

Mathematical Framework

The formal justification of the Second Derivative Test rests on the local Taylor expansion of f about the critical point c. Because f′(c) = 0, the first-order term vanishes, and the behavior of f near c is dominated by the second-order term involving f″(c). This is why the sign of f″(c) alone suffices to classify the critical point, provided f″(c) ≠ 0.

TAYLOR EXPANSION AT A CRITICAL POINT

f(c + h) = f(c) + f′(c)·h + ½·f″(c)·h² + O(h³)

Since f′(c) = 0, this simplifies to f(c + h) ≈ f(c) + ½·f″(c)·h². The sign of f″(c) determines whether f(c + h) is greater than or less than f(c) for small h.

SECOND DERIVATIVE TEST — LOCAL MINIMUM

If f′(c) = 0 and f″(c) > 0, then f has a local minimum at x = c.

When f″(c) > 0, the term ½·f″(c)·h² > 0 for all small h ≠ 0, so f(c + h) > f(c). The function value at c is less than all nearby values—a local minimum.

SECOND DERIVATIVE TEST — LOCAL MAXIMUM

If f′(c) = 0 and f″(c) < 0, then f has a local maximum at x = c.

When f″(c) < 0, the term ½·f″(c)·h² < 0 for all small h ≠ 0, so f(c + h) < f(c). The function value at c exceeds all nearby values—a local maximum.

INCONCLUSIVE CASE

If f′(c) = 0 and f″(c) = 0, the Second Derivative Test is inconclusive.

Consider f(x) = x⁴ (local min at 0), f(x) = −x⁴ (local max at 0), and f(x) = x³ (inflection at 0). All have f′(0) = 0 and f″(0) = 0, yet exhibit three different behaviors. The First Derivative Test or higher-order analysis is required.

💡

AP Exam Tip

On the AP Calculus BC exam, always state the conditions you are checking: "f′(c) = 0 and f″(c) > 0, so by the Second Derivative Test, f has a local minimum at x = c." Omitting the justification—even if your answer is correct—can cost rubric points on free-response questions.

SECTION 5

Classification Flowchart & Decision Logic

When confronted with a function whose extrema you must classify, the decision process follows a clear algorithmic path. First, compute f′(x) and solve f′(x) = 0 to locate all critical points where the derivative vanishes. Next, compute f″(x) and evaluate it at each critical point. The sign of f″ at that point determines the classification. The flowchart below captures this decision tree, including the fallback path when the test is inconclusive.

Decision flowchart for the Second Derivative Test. After finding critical points by setting f′(c) = 0, evaluate f″(c). A negative value yields a local maximum (pink), a positive value yields a local minimum (cyan), and zero means the test is inconclusive (amber), requiring the First Derivative Test or higher-order analysis.

Condition	f″(c)	Conclusion	Geometric Meaning
f′(c) = 0	> 0	Local minimum	Curve is concave up — shaped like a bowl
f′(c) = 0	< 0	Local maximum	Curve is concave down — shaped like a dome
f′(c) = 0	= 0	Inconclusive	Curvature vanishes — could be max, min, or inflection

Summary of Second Derivative Test outcomes

SECTION 6

Worked Example

Let us apply the Second Derivative Test to classify the critical points of a polynomial function that appears frequently in AP Calculus BC problems. Consider the function f(x) = 2x³ − 9x² + 12x − 4. We will find all critical points, evaluate f″ at each, and classify them.

Classify the extrema of f(x) = 2x³ − 9x² + 12x − 4

1
Step 1 — Compute the first derivative

Differentiate f(x) = 2x³ − 9x² + 12x − 4 using the power rule. f′(x) = 6x² − 18x + 12. Factor out the common factor of 6: f′(x) = 6(x² − 3x + 2) = 6(x − 1)(x − 2).

f′(x) = 6(x − 1)(x − 2)
2
Step 2 — Find the critical points

Set f′(x) = 0 and solve. 6(x − 1)(x − 2) = 0 gives x = 1 and x = 2. Since f′ is defined for all real x, these are the only critical points.

Critical points: x = 1 and x = 2
3
Step 3 — Compute the second derivative

Differentiate f′(x) = 6x² − 18x + 12 to obtain f″(x) = 12x − 18.

f″(x) = 12x − 18
4
Step 4 — Evaluate f″ at each critical point

At x = 1: f″(1) = 12(1) − 18 = −6. Since f″(1) < 0, the graph is concave down at x = 1, so f has a local maximum at x = 1. At x = 2: f″(2) = 12(2) − 18 = 6. Since f″(2) > 0, the graph is concave up at x = 2, so f has a local minimum at x = 2.

f″(1) = −6 → local max; f″(2) = 6 → local min
5
Step 5 — State the extrema values

Compute the function values: f(1) = 2(1)³ − 9(1)² + 12(1) − 4 = 2 − 9 + 12 − 4 = 1. f(2) = 2(2)³ − 9(2)² + 12(2) − 4 = 16 − 36 + 24 − 4 = 0. Therefore, f has a local maximum of 1 at x = 1 and a local minimum of 0 at x = 2.

Local maximum: (1, 1); Local minimum: (2, 0)

SECTION 7

First Derivative Test vs. Second Derivative Test

Both the First Derivative Test and the Second Derivative Test serve the same purpose—classifying critical points as local maxima, local minima, or neither—but they do so through different mechanisms. Understanding when each test is preferable is essential for both exam efficiency and deeper conceptual understanding. The following comparison highlights their strengths and limitations.

Feature	First Derivative Test	Second Derivative Test
What you compute	Sign of f′ on intervals flanking c	Value of f″(c) at the single point c
Handles f′(c) DNE	Yes — works at cusps and corners	No — requires f″(c) to exist
Inconclusive case	Never inconclusive (always classifies or identifies neither)	Inconclusive when f″(c) = 0
Computational effort	Must test points in two intervals and determine signs	Single evaluation of f″(c) — often faster
Best used when	f″ is hard to compute or f″(c) = 0	f″ is easy to compute and f″(c) ≠ 0

Comparison of the First and Second Derivative Tests for classifying critical points

✦

KEY TAKEAWAY

Think of the Second Derivative Test as a quick diagnostic scan and the First Derivative Test as a comprehensive MRI. The scan (SDT) is fast and usually conclusive, but occasionally the results come back "unclear" (when f″(c) = 0), and you need the more thorough procedure (FDT) to get a definitive answer. On the AP exam, attempt the Second Derivative Test first for efficiency, and fall back to the First Derivative Test only when needed.

SECTION 8

Connection to Advanced Theory

The Second Derivative Test for single-variable functions is the gateway to more powerful optimization machinery that you will encounter in multivariable calculus and beyond. In Calculus III, the analogous test for functions of two variables f(x, y) uses the Hessian matrix—a matrix of second partial derivatives—to classify critical points. The determinant of the Hessian and the sign of f_xx play roles analogous to f″(c) in the single-variable case.

Feature	Single-Variable (Calc BC)	Multivariable (Calc III)
Critical point condition	f′(c) = 0	∇f(a, b) = 0 (both partials vanish)
Classification tool	Sign of f″(c)	Determinant of Hessian: D = f_xx·f_yy − (f_xy)²
Local min condition	f″(c) > 0	D > 0 and f_xx > 0
Local max condition	f″(c) < 0	D > 0 and f_xx < 0
New phenomenon	N/A	D < 0 → saddle point (no single-variable analog)

Second Derivative Test: single-variable vs. multivariable extensions

In optimization theory and machine learning, second-derivative information is used ubiquitously. Newton's method for optimization uses the Hessian to find critical points of objective functions, and the second-order conditions determine whether the algorithm has converged to a minimum rather than a maximum or saddle point. Mastering the single-variable Second Derivative Test now gives you the conceptual foundation for these advanced applications, where the principle is identical—it is the sign of the "curvature" at a stationary point that determines its nature—but the computational machinery scales to arbitrary dimensions.

SECTION 9

Practice Problems

Question 1: CONCEPTUAL

Suppose f is a twice-differentiable function with f′(3) = 0 and f″(3) = −7. Which of the following is true?

A. f has a local minimum at x = 3.
B. f has a local maximum at x = 3.
C. f has an inflection point at x = 3.
D. f is concave up at x = 3.
E. The Second Derivative Test is inconclusive at x = 3.

Show answer (B)

The correct answer is B. Since f′(3) = 0 and f″(3) = −7 < 0, the Second Derivative Test concludes that f has a local maximum at x = 3. Choice A is wrong because a local minimum requires f″(c) > 0. Choice C is wrong because an inflection point is where concavity changes, and we only know f″(3) < 0; moreover, f′(3) = 0 with f″(3) ≠ 0 means this is an extremum, not an inflection point. Choice D is wrong because f″(3) < 0 means concave down, not concave up. Choice E is wrong because the test is only inconclusive when f″(c) = 0, and here f″(3) = −7 ≠ 0.

Question 2: BASIC CALCULATION

Let g(x) = x⁴ − 8x² + 3. Using the Second Derivative Test, classify all critical points where the test is conclusive. At which x-value does g have a local maximum?

A. x = 0 only
B. x = 2 only
C. x = −2 and x = 2
D. x = 0, x = −2, and x = 2
E. The Second Derivative Test is inconclusive at all critical points.

Show answer (A)

The correct answer is A. First, g′(x) = 4x³ − 16x = 4x(x² − 4) = 4x(x − 2)(x + 2), so the critical points are x = 0, x = 2, and x = −2. Next, g″(x) = 12x² − 16. Evaluating: g″(0) = −16 < 0, so x = 0 is a local maximum. g″(2) = 12(4) − 16 = 32 > 0, so x = 2 is a local minimum. g″(−2) = 12(4) − 16 = 32 > 0, so x = −2 is also a local minimum. The only local maximum occurs at x = 0. Choice B names a local minimum. Choice C names both local minima. Choice D lists all critical points but only x = 0 is a local max. Choice E is false since f″ is nonzero at all three critical points.

Question 3: INTERMEDIATE

Let h(x) =

xe^{−x}

. Find all critical points and use the Second Derivative Test to classify each. Which of the following correctly identifies the local extremum?

A. Local minimum at x = 1 with h(1) = e⁻¹
B. Local maximum at x = 1 with h(1) = e⁻¹
C. Local maximum at x = −1 with h(−1) = −e
D. Local minimum at x = 0 with h(0) = 0
E. The Second Derivative Test is inconclusive at x = 1.

Show answer (B)

The correct answer is B. Using the product rule, h′(x) =

e^{−x}

+ x(−

e^{−x}

) =

e^{−x}

(1 − x). Setting h′(x) = 0: since

e^{−x}

> 0 for all x, we need 1 − x = 0, so x = 1 is the only critical point. For the second derivative: h″(x) = −

e^{−x}

(1 − x) +

e^{−x}

(−1) =

e^{−x}

(x − 2). At x = 1: h″(1) = e⁻¹(1 − 2) = −e⁻¹ < 0, so x = 1 is a local maximum. h(1) = 1·e⁻¹ = e⁻¹ ≈ 0.368. Choice A incorrectly labels it as a minimum. Choice C invents a critical point at x = −1 which does not satisfy h′(−1) = 0. Choice D claims x = 0 is a critical point, but h′(0) = e⁰(1) = 1 ≠ 0. Choice E is wrong since h″(1) = −e⁻¹ ≠ 0.

Question 4: APPLIED

4 pts

A manufacturer determines that the profit P (in thousands of dollars) from producing x hundred units of a product is given by P(x) = −2x³ + 15x² − 36x + 40 for x > 0. Use the Second Derivative Test to determine the production level that yields a local maximum profit. (a) Find all critical points of P(x) for x > 0. (b) Apply the Second Derivative Test to classify each critical point. (c) State the production level (in units) that maximizes local profit and compute that profit. (d) Explain why the Second Derivative Test is sufficient here and the First Derivative Test is not needed.

Show answer

Point 1 (Finding critical points): P′(x) = −6x² + 30x − 36 = −6(x² − 5x + 6) = −6(x − 2)(x − 3). Setting P′(x) = 0 gives x = 2 and x = 3, both in the domain x > 0. Point 2 (Second Derivative Test): P″(x) = −12x + 30. At x = 2: P″(2) = −24 + 30 = 6 > 0, so x = 2 is a local minimum. At x = 3: P″(3) = −36 + 30 = −6 < 0, so x = 3 is a local maximum. Point 3 (Computing the answer): Since x is in hundreds of units, x = 3 corresponds to 300 units. P(3) = −2(27) + 15(9) − 36(3) + 40 = −54 + 135 − 108 + 40 = 13. The local maximum profit is $13,000 when 300 units are produced. Point 4 (Justification of sufficiency): The Second Derivative Test is sufficient because f″ is nonzero at both critical points (f″(2) = 6 ≠ 0 and f″(3) = −6 ≠ 0), so the test conclusively classifies each as a local min or local max. Since there is no inconclusive case, the First Derivative Test is not needed as a fallback.

Question 5: CRITICAL THINKING

3 pts

Consider the family of functions

f_n

(x) =

x^{2n}

for positive integers n ≥ 1. (a) Show that x = 0 is a critical point for every

f_n

. (b) Determine for which values of n the Second Derivative Test is conclusive at x = 0, and for which it is inconclusive. (c) For the inconclusive case(s), use the First Derivative Test or another argument to classify x = 0.

Show answer

Point 1 (Critical point verification):

f_n

′(x) = 2n·

x^{2n−1}

. For all n ≥ 1,

f_n

′(0) = 0, so x = 0 is a critical point. Point 2 (Conclusive vs. inconclusive):

f_n

″(x) = 2n(2n−1)·

x^{2n−2}

. When n = 1:

f_1

″(x) = 2, so

f_1

″(0) = 2 > 0. The SDT is conclusive and yields a local minimum. When n ≥ 2:

f_n

″(0) = 2n(2n−1)·

0^{2n−2}

= 0 (since 2n − 2 ≥ 2 > 0). The SDT is inconclusive for all n ≥ 2. Point 3 (Resolving the inconclusive cases): For n ≥ 2,

f_n

(x) =

x^{2n}

≥ 0 for all x, with equality only at x = 0. Since

f_n

(0) = 0 ≤

f_n

(x) for all x near 0, x = 0 is a local (in fact, global) minimum. Alternatively, by the First Derivative Test: for x < 0,

f_n

′(x) = 2n·

x^{2n−1}

< 0 (since 2n − 1 is odd), and for x > 0,

f_n

′(x) > 0. The sign change from negative to positive confirms a local minimum. This example illustrates that the Second Derivative Test's inconclusive verdict does not mean the point fails to be an extremum—it merely means the test lacks the power to determine the classification.

SUMMARY

Summary

The Second Derivative Test provides a streamlined method for classifying critical points of a twice-differentiable function. After locating all values c where f′(c) = 0, you evaluate f″(c). If f″(c) > 0, the curve is concave up and f has a local minimum; if f″(c) < 0, the curve is concave down and f has a local maximum; if f″(c) = 0, the test is inconclusive and you must fall back on the First Derivative Test.

The test's power comes from the Taylor expansion at the critical point: with the first-order term vanishing, the second-order term ½·f″(c)·h² dominates the behavior of f near c. This single-variable framework extends naturally to the Hessian matrix in multivariable calculus. On the AP exam, remember to state both conditions—f′(c) = 0 and the sign of f″(c)—and explicitly name the test for full justification credit.