Alternating Series Test for Convergence - AP Calculus BC

SECTION 1

Historical Context & Motivation

The study of infinite series has been central to mathematics since antiquity, but the rigorous treatment of series whose terms alternate in sign represents a particularly elegant chapter in the history of analysis. Early mathematicians recognized that some divergent-looking series could in fact converge, provided their terms oscillated between positive and negative values in a controlled manner. The Alternating Series Test (also called the Leibniz criterion) codifies the precise conditions under which such series converge, and its development reflects a broader movement toward placing calculus on firm logical foundations.

1668 • Mercator's Series

Nicolaus Mercator published the series ln(1 + x) = x − x²/2 + x³/3 − ⋯, one of the earliest recognized alternating series, providing a concrete example of convergent sign-alternating sums.

1682 • Leibniz and the Alternating Harmonic Series

Gottfried Wilhelm Leibniz demonstrated that the series 1 − 1/2 + 1/3 − 1/4 + ⋯ converges to ln 2, establishing the first clear instance of the alternating series test in action and providing an informal argument for convergence.

1821 • Cauchy's Rigorous Foundation

Augustin-Louis Cauchy, in his Cours d'Analyse, gave the first rigorous proof of the alternating series test using modern epsilon-based arguments and partial sum analysis, anchoring it within the framework of real analysis.

1854 • Riemann's Rearrangement Theorem

Bernhard Riemann proved that the terms of a conditionally convergent alternating series can be rearranged to converge to any desired real number (or diverge), underscoring the delicate nature of alternating series convergence.

The central question that motivated these developments remains directly relevant for AP Calculus BC: given an infinite series whose terms alternate in sign, how can we determine whether the partial sums settle toward a finite value? The Alternating Series Test provides a clean, two-condition answer that is both easy to apply and surprisingly powerful, covering many series that other convergence tests cannot handle as efficiently.

SECTION 2

Core Principles & Definitions

An alternating series is any infinite series whose successive terms alternate between positive and negative values. Such a series can be written in one of two standard forms: ∑(−1)ⁿ bₙ or ∑(−1)ⁿ⁺¹ bₙ, where bₙ > 0 for all n. The factor (−1)ⁿ handles the sign alternation, while the sequence {bₙ} captures the magnitude of each term. The Alternating Series Test (AST) specifies precisely when such series converge, and understanding its two conditions—along with their geometric interpretation—is essential for the AP exam.

Alternating Form

The series must genuinely alternate in sign. It takes the form ∑(−1)ⁿ bₙ or ∑(−1)ⁿ⁺¹ bₙ with bₙ > 0. Every consecutive pair of terms has opposite signs.

Decreasing Condition

The absolute values of the terms must be eventually decreasing: bₙ₊₁ ≤ bₙ for all n beyond some index N. This ensures each successive correction overshoots less than the previous one.

Limit-to-Zero Condition

The limit of the magnitudes must be zero: lim(n→∞) bₙ = 0. If this fails, the nth-term test for divergence already guarantees the series diverges, so this condition is necessary for any convergence.

Conditional vs. Absolute Convergence

The AST proves convergence but says nothing about absolute convergence. A series may be conditionally convergent (converges, but ∑|aₙ| diverges) or absolutely convergent. Always check absolute convergence separately when needed.

✦

KEY TAKEAWAY

Think of an alternating series like a pendulum that swings a little less each time. The first term swings the partial sum to the right, the second swings it back to the left (but not as far), the third swings right again (even less), and so on. As long as each swing is smaller than the last and the swing amplitude approaches zero, the pendulum settles at a definite resting point—the series converges. If the swings don't shrink to zero, the pendulum never settles, and the series diverges.

SECTION 3

Visual Explanation: Partial Sums of an Alternating Series

The diagram plots the partial sums S₁ through S₉ of the alternating harmonic series. Odd partial sums (pink dots) approach ln 2 from above, while even partial sums (cyan dots) approach from below. The oscillation narrows with each term, visually confirming convergence to the dashed green line at ln 2 ≈ 0.6931.

The oscillating staircase pattern visible in the diagram is the geometric hallmark of any convergent alternating series. Notice how each odd partial sum (S₁, S₃, S₅, …) overshoots the limit from above, while each even partial sum (S₂, S₄, S₆, …) undershoots from below. Because each successive term has a smaller absolute value than the one before it, the overshoots and undershoots diminish monotonically, and the two subsequences squeeze toward the same value. This "squeezing" mechanism is precisely what the Alternating Series Test guarantees: when the decreasing and limit-to-zero conditions hold, the partial sums are trapped between two converging bounds, forcing the series to converge.

SECTION 4

Mathematical Framework

The formal statement of the Alternating Series Test brings together the sign-alternation structure and the two convergence conditions into a single theorem. Understanding the formal notation is important for both applying the test correctly and writing rigorous justifications on the AP free-response section.

ALTERNATING SERIES TEST (LEIBNIZ CRITERION)

If ∑(−1)ⁿ bₙ (or ∑(−1)ⁿ⁺¹ bₙ) with bₙ > 0 satisfies: (1) bₙ₊₁ ≤ bₙ for all n ≥ N, and (2) lim(n→∞) bₙ = 0, then the series converges.

bₙ represents the absolute value (magnitude) of the nth term. Condition (1) requires the magnitudes to be eventually non-increasing. Condition (2) is the necessary condition for any series to converge (nth-term test), applied here to the magnitudes.

ALTERNATING SERIES ESTIMATION THEOREM

|S − Sₙ| ≤ bₙ₊₁

S is the exact sum of the convergent alternating series, and Sₙ is the nth partial sum. The absolute error is bounded by the magnitude of the first omitted term. This bound is a powerful tool on the AP exam for determining how many terms are needed for a given accuracy.

It is worth understanding why the test works at a proof-sketch level. Consider the subsequence of even partial sums: S₂ₙ = (b₁ − b₂) + (b₃ − b₄) + ⋯ + (b₂ₙ₋₁ − b₂ₙ). Because bₖ ≥ bₖ₊₁, each parenthesized pair is non-negative, so {S₂ₙ} is non-decreasing. Alternatively, S₂ₙ = b₁ − (b₂ − b₃) − (b₄ − b₅) − ⋯ − b₂ₙ, which shows S₂ₙ ≤ b₁, so {S₂ₙ} is bounded above. A bounded, non-decreasing sequence converges by the Monotone Convergence Theorem. Similarly, the odd partial sums form a non-increasing sequence bounded below. Since S₂ₙ₊₁ − S₂ₙ = b₂ₙ₊₁ → 0, the two subsequences share the same limit, and the full sequence of partial sums converges.

CHECKING THE DECREASING CONDITION

bₙ₊₁ ≤ bₙ ⟺ bₙ₊₁/bₙ ≤ 1 ⟺ f'(x) ≤ 0 where f(x) = bₓ

Three equivalent methods to verify the decreasing condition: (a) direct comparison of consecutive terms, (b) computing the ratio bₙ₊₁/bₙ, or (c) defining a continuous function f(x) = bₓ and showing its derivative is non-positive for sufficiently large x.

SECTION 5

Conditional vs. Absolute Convergence

Passing the Alternating Series Test tells you a series converges, but it does not tell you how strongly it converges. The distinction between absolute convergence and conditional convergence is crucial on the AP exam because it determines the robustness of the convergence and affects the behavior of the series under rearrangement. A series ∑aₙ converges absolutely if ∑|aₙ| also converges; it converges conditionally if ∑aₙ converges but ∑|aₙ| diverges. The alternating harmonic series is the canonical example of conditional convergence: ∑(−1)ⁿ⁺¹/n converges by the AST, but the harmonic series ∑1/n diverges.

This flowchart outlines the decision process for classifying an alternating series. First test for absolute convergence; if that fails, apply the AST to determine conditional convergence or divergence.

Example Series	∑\|aₙ\| Behavior	Classification
∑(−1)ⁿ / n!	∑1/n! converges (ratio test)	Absolutely convergent
∑(−1)ⁿ⁺¹ / n	∑1/n diverges (harmonic)	Conditionally convergent
∑(−1)ⁿ / n²	∑1/n² converges (p-series, p = 2)	Absolutely convergent
∑(−1)ⁿ / √n	∑1/√n diverges (p-series, p = 1/2)	Conditionally convergent

Common alternating series and their convergence classification

SECTION 6

Worked Example

Determine whether ∑(−1)ⁿ⁺¹ · n/(n² + 1) converges or diverges, and if it converges, classify it as absolutely or conditionally convergent.

1
Step 1 — Identify the Alternating Structure

The series is ∑ from n = 1 to ∞ of (−1)ⁿ⁺¹ · n/(n² + 1). The factor (−1)ⁿ⁺¹ ensures sign alternation. We define bₙ = n/(n² + 1), which is positive for all n ≥ 1.

bₙ = n/(n² + 1) > 0 ✓
2
Step 2 — Verify the Limit-to-Zero Condition

Compute lim(n→∞) n/(n² + 1). Dividing numerator and denominator by n² gives lim(n→∞) (1/n)/(1 + 1/n²) = 0/1 = 0. The limit condition is satisfied.

lim(n→∞) bₙ = 0 ✓
3
Step 3 — Verify the Decreasing Condition

Let f(x) = x/(x² + 1). Compute f'(x) using the quotient rule: f'(x) = (x² + 1 − x · 2x)/(x² + 1)² = (1 − x²)/(x² + 1)². For x ≥ 1, the numerator 1 − x² ≤ 0, so f'(x) ≤ 0. Therefore f is non-increasing on [1, ∞), which means bₙ₊₁ ≤ bₙ for all n ≥ 1.

bₙ₊₁ ≤ bₙ for n ≥ 1 ✓
4
Step 4 — Apply the AST

Both conditions of the Alternating Series Test are satisfied. Therefore, the series ∑(−1)ⁿ⁺¹ · n/(n² + 1) converges.

The series converges by the AST.
5
Step 5 — Classify: Absolute or Conditional Convergence

Check whether ∑ n/(n² + 1) converges. By limit comparison with the harmonic series ∑1/n: lim(n→∞) [n/(n² + 1)] / (1/n) = lim(n→∞) n²/(n² + 1) = 1 > 0 and finite. Since ∑1/n diverges, the limit comparison test tells us ∑ n/(n² + 1) also diverges. The series does not converge absolutely.

The series is conditionally convergent.

SECTION 7

AST vs. Other Convergence Tests

One of the most important skills for the AP Calculus BC exam is choosing the right convergence test for a given series. The Alternating Series Test occupies a specific niche—it applies only to series with alternating signs—and understanding how it compares to other tests will help you recognize when to reach for it and when to use something else.

Convergence Test	Best Used When	Key Limitation
Alternating Series Test	Series has form ∑(−1)ⁿ bₙ with positive bₙ that decrease to 0	Only proves convergence, not absolute convergence; only applies to alternating series
Ratio Test	Terms involve factorials, exponentials, or powers of n	Inconclusive when the ratio limit equals 1
Integral Test	f(n) = aₙ is positive, continuous, and decreasing; antiderivative is findable	Requires a positive series; can be computationally heavy
Direct / Limit Comparison	Terms resemble a known benchmark series (p-series, geometric)	Requires positive terms; choosing the comparator can be tricky
p-Series Test	Series has the exact form ∑1/nᵖ	Only applies to this specific form; often used as a comparison benchmark

Comparison of convergence tests commonly tested on AP Calculus BC

✦

KEY TAKEAWAY

Think of convergence tests as diagnostic tools in an engineer's toolkit. The AST is like a specialized gauge designed for one specific type of measurement (alternating series)—it gives a quick, reliable reading when applicable but cannot measure anything else. The Ratio Test and Comparison Tests are more like general-purpose multimeters: they work in broader contexts but sometimes return an inconclusive reading. On the AP exam, your first move with any alternating series should be to check whether the AST applies; only if it fails or if the question asks about absolute convergence should you pivot to another test.

SECTION 8

Connections to Power Series & Taylor Series

The Alternating Series Test and its companion Estimation Theorem are not merely standalone tools—they connect directly to some of the most powerful ideas in AP Calculus BC, particularly power series and Taylor/Maclaurin series. When you evaluate a Taylor series at a specific value of x, the resulting numerical series is often alternating. For example, the Maclaurin series for eˣ evaluated at x = −1 produces the alternating series ∑(−1)ⁿ/n!, which converges to e⁻¹. The AST then guarantees convergence, and the Estimation Theorem lets you determine how many terms are needed to approximate the value to a specified accuracy.

AST Concept	Advanced Extension
Convergence of alternating series	Determines convergence at endpoints of power series intervals; essential for testing whether the interval of convergence includes its boundary
Alternating Series Estimation Theorem	Bounds the error when approximating a function by a Taylor polynomial; commonly tested in AP FRQ (e.g., "find the least n such that the error is less than 0.001")
Conditional vs. absolute convergence	Determines whether a power series converges absolutely (safely rearrangeable) or conditionally at an endpoint, affecting the nature of convergence on the boundary
Decreasing magnitudes to zero	Connects to Riemann's Rearrangement Theorem: conditionally convergent series can be rearranged to sum to any value, a result studied in real analysis courses

How AST concepts extend to advanced topics

In practice, the most common advanced application of the AST on the AP exam involves determining convergence at the endpoints of a power series' interval of convergence. After using the Ratio Test to find the open interval (−R, R), you substitute x = −R and x = R into the series and check each endpoint individually. Frequently, one or both endpoints yield an alternating series, and the AST becomes the decisive tool. This endpoint analysis is a recurring theme in both multiple-choice and free-response questions, so mastering the AST directly strengthens your performance across the entire series unit.

SECTION 9

Practice Problems

Question 1: CONCEPTUAL

Which of the following is a necessary condition for the Alternating Series Test to guarantee convergence of ∑(−1)ⁿ bₙ?

A. bₙ must be positive and increasing for all n.
B. ∑ bₙ must converge.
C. lim(n→∞) bₙ = 0 and bₙ is eventually non-increasing.
D. The series must converge absolutely.
E. bₙ must be a geometric sequence with |r| < 1.

Show answer (C)

The correct answer is C. The Alternating Series Test requires two conditions: (1) the sequence bₙ is eventually non-increasing, and (2) lim(n→∞) bₙ = 0. Choice A is incorrect because the test requires bₙ to be decreasing, not increasing. Choice B is incorrect because convergence of ∑ bₙ would imply absolute convergence, which is sufficient but not necessary for the AST. Choice D is incorrect because the AST can prove convergence even when absolute convergence fails (conditional convergence). Choice E is incorrect because the AST applies to any alternating series meeting the two conditions, not just geometric series.

Question 2: BASIC CALCULATION

Determine whether the series ∑ from n = 1 to ∞ of (−1)ⁿ⁺¹ · 1/(2n − 1) converges.

A. The series diverges because lim(n→∞) bₙ ≠ 0.
B. The series converges absolutely by the p-series test.
C. The series converges conditionally by the Alternating Series Test.
D. The series diverges because bₙ is not decreasing.
E. The Alternating Series Test is inconclusive for this series.

Show answer (C)

The correct answer is C. Here bₙ = 1/(2n − 1). First, bₙ > 0 for all n ≥ 1. Second, bₙ₊₁ = 1/(2n + 1) < 1/(2n − 1) = bₙ, so the sequence is decreasing. Third, lim(n→∞) 1/(2n − 1) = 0. Both conditions of the AST are met, so the series converges. However, ∑ 1/(2n − 1) diverges by limit comparison with ∑ 1/n (the limit of [1/(2n−1)]/(1/n) = n/(2n−1) → 1/2), so the convergence is conditional, not absolute. Choice A is wrong because the limit is in fact 0. Choice B is wrong because ∑ 1/(2n−1) diverges. Choice D is wrong because bₙ is indeed decreasing. Choice E is wrong because the test gives a definitive result.

Question 3: INTERMEDIATE

Consider the series ∑ from n = 1 to ∞ of (−1)ⁿ · n/(n + 3). Does the series converge or diverge?

A. Converges by the AST because the terms alternate in sign.
B. Converges conditionally because bₙ is eventually decreasing.
C. Converges absolutely by the Ratio Test.
D. Diverges by the nth-Term Test for Divergence.
E. Diverges by the Integral Test.

Show answer (D)

The correct answer is D. Here bₙ = n/(n + 3), and lim(n→∞) n/(n + 3) = 1 ≠ 0. Since the magnitudes of the terms do not approach zero, the general term (−1)ⁿ · n/(n + 3) does not approach zero either; it oscillates between values near −1 and +1. By the nth-Term Test for Divergence, if lim(n→∞) aₙ ≠ 0, the series diverges. Choice A is wrong because alternation alone is insufficient—the limit condition must also hold. Choice B is wrong because even though bₙ is not decreasing (it is increasing), the decisive reason for divergence is the failed limit condition. Choice C is wrong because the Ratio Test applied to |aₙ| = n/(n+3) gives limit 1, which is inconclusive, and furthermore the series diverges. Choice E, while the Integral Test could show ∑ n/(n+3) diverges, is not the most direct argument, and the series diverges by the more fundamental nth-Term Test.

Question 4: APPLIED

4 pts

The Maclaurin series for ln(1 + x) is ∑ from n = 1 to ∞ of (−1)ⁿ⁺¹ · xⁿ/n, valid for −1 < x ≤ 1. Using this series evaluated at x = 1, determine the minimum number of terms needed so that the partial sum approximates ln 2 with an error less than 0.05. Show your work and justify your answer using the Alternating Series Estimation Theorem.

Show answer

Point 1 (1 pt): Correctly identifies the series at x = 1 as ∑(−1)ⁿ⁺¹/n = 1 − 1/2 + 1/3 − 1/4 + ⋯ and states that this series converges to ln 2 by the AST. Point 2 (1 pt): States the Alternating Series Estimation Theorem: the error |S − Sₙ| ≤ bₙ₊₁ where bₙ₊₁ = 1/(n + 1) is the magnitude of the first omitted term. Point 3 (1 pt): Sets up the inequality 1/(n + 1) < 0.05, which yields n + 1 > 20, so n > 19, meaning n ≥ 20. Point 4 (1 pt): Concludes that the minimum number of terms is n = 20, and verifies: S₂₀ uses terms through 1/20, and the first omitted term has magnitude 1/21 ≈ 0.0476 < 0.05, while using n = 19 would give first omitted term 1/20 = 0.05, which does not satisfy the strict inequality < 0.05. Therefore at least 20 terms are required.

Question 5: CRITICAL THINKING

3 pts

A student claims that the series ∑ from n = 1 to ∞ of (−1)ⁿ · sin(1/n) diverges because sin(1/n) is not a decreasing function near n = 1. Evaluate this claim. Determine whether the series converges or diverges, classify the convergence type if applicable, and address the student's reasoning.

Show answer

Point 1 (1 pt): The student's claim is incorrect. The AST requires bₙ to be eventually non-increasing, not necessarily decreasing from n = 1. Let bₙ = sin(1/n). For the function f(x) = sin(1/x), f'(x) = −cos(1/x)/x², which is negative whenever cos(1/x) > 0. For x ≥ 1, 0 < 1/x ≤ 1 < π/2, so cos(1/x) > 0, and f is indeed decreasing on [1, ∞). Thus bₙ is decreasing for all n ≥ 1, and the student's observation is factually wrong. Point 2 (1 pt): Verify the limit condition: lim(n→∞) sin(1/n) = sin(0) = 0. Both AST conditions hold, so the series ∑(−1)ⁿ sin(1/n) converges. Point 3 (1 pt): To classify convergence, consider ∑ sin(1/n). Since lim(n→∞) sin(1/n)/(1/n) = 1 by the small-angle limit, and ∑ 1/n diverges, the Limit Comparison Test shows ∑ sin(1/n) diverges. Therefore the series converges conditionally, not absolutely.

SUMMARY

Summary & Key Concepts

Summary

The Alternating Series Test provides a straightforward criterion for determining the convergence of series whose terms alternate in sign. If the sequence of absolute values {bₙ} is eventually non-increasing and lim(n→∞) bₙ = 0, the alternating series converges. The partial sums oscillate around the true sum with ever-decreasing amplitude, producing the characteristic "squeeze" toward a finite limit. The companion Alternating Series Estimation Theorem bounds the error by the magnitude of the first omitted term: |S − Sₙ| ≤ bₙ₊₁.

Crucially, passing the AST alone establishes convergence but not absolute convergence. You must separately test ∑|aₙ| to distinguish between conditional convergence (where ∑|aₙ| diverges) and absolute convergence (where ∑|aₙ| also converges). On the AP Calculus BC exam, the AST appears in stand-alone convergence questions, in endpoint analysis of power series, and in Taylor polynomial error estimation problems—making it one of the most versatile and frequently tested tools in the infinite series unit.