AP Calculus: Alternating Series and Alternating Series Error Bound

Alternating Series and Alternating Series Error Bound

Concept Summary

• Alternating Series: A series of the form (\sum (-1)^{n+1} a_n) or (\sum (-1)^n a_n) where (a_n > 0) for all (n); converges if terms decrease in magnitude and approach zero.
• Alternating Series Test (AST): If (a_{n+1} \leq a_n) for all (n) and (\lim_{n \to \infty} a_n = 0), the series converges.
• Alternating Series Error Bound: The absolute error (|S - S_n|) when approximating the sum (S) by the (n)th partial sum (S_n) is less than or equal to the first omitted term (a_{n+1}).
• Absolute vs. Conditional Convergence: An alternating series may converge conditionally (converges but (\sum |a_n|) diverges) or absolutely (both converge).
• Significance of Error Bound: Provides a simple, calculable guarantee for approximation accuracy without needing the exact sum.

Core Questions

WHAT (definitional)

Q: What is an alternating series? A: A series where terms alternate in sign, typically written as (\sum (-1)^{n+1} a_n) with (a_n > 0). Trap/Clarification: The series (\sum (-1)^n a_n) is also alternating, but the exponent’s parity shifts the starting sign (e.g., (n=0) starts negative).

Q: What is the Alternating Series Error Bound? A: The maximum error when approximating the sum (S) by (S_n) is (|S - S_n| \leq a_{n+1}). Trap/Clarification: The bound is only valid if the series satisfies the AST conditions (decreasing (a_n), (\lim a_n = 0)).

WHY (causal/explanatory)

Q: Why does the Alternating Series Test require (a_{n+1} \leq a_n)? A: Ensures terms oscillate with decreasing magnitude, preventing divergence due to unbounded oscillations (e.g., ((-1)^n n)). Trap/Clarification: The test fails if (a_n) is not eventually decreasing (e.g., (a_n = 1/n) for (n) odd, (a_n = 1/n^2) for (n) even).

Q: Why is the error bound (a_{n+1}) useful? A: Provides a computable upper bound for the error without knowing the exact sum (S), critical for approximations. Trap/Clarification: The bound is not the actual error—it’s a worst-case guarantee (actual error may be smaller).

HOW (process/application)

Q: How do you apply the Alternating Series Test? A: Verify (1) (a_{n+1} \leq a_n) for all (n \geq N) (eventually decreasing), and (2) (\lim_{n \to \infty} a_n = 0). Trap/Clarification: The test is not for divergence—failure of AST does not imply divergence (e.g., (\sum (-1)^n / \sqrt{n}) converges despite (1/\sqrt{n}) not decreasing for all (n)).

Q: How is the error bound calculated for an alternating series? A: For (S \approx S_n), the error (|S - S_n| \leq a_{n+1}). To find (n) for a desired error (\epsilon), solve (a_{n+1} \leq \epsilon). Trap/Clarification: The bound only applies to series satisfying AST—check conditions first!

CAN (conditions/possibilities)

Q: Can an alternating series converge if (\lim_{n \to \infty} a_n \neq 0)? A: No; divergence is guaranteed by the Divergence Test if (\lim a_n \neq 0). Trap/Clarification: The converse is false: (\lim a_n = 0) does not guarantee convergence (e.g., harmonic series).

Q: Under what conditions is the error bound (|S - S_n| \leq a_{n+1}) exact? A: Never; the bound is always an overestimate (actual error is (\leq a_{n+1}) but often smaller). Trap/Clarification: The error alternates in sign, so (S) lies between (S_n) and (S_{n+1}).

Quick Facts & Traps

• Fact: AST is a one-way test: Convergence (\implies) AST conditions met, but failure of AST (\not\implies) divergence.
• Trap: Ignoring "eventually decreasing": A single term violating (a_{n+1} \leq a_n) (e.g., (a_1 = 1), (a_2 = 2)) can invalidate AST. -Reality: Check for (n \geq N) where the sequence becomes decreasing.
• Fact: Error bound is tight for the first term: For (n=0), (|S - S_0| = a_1) (exact error).
• Trap: Assuming absolute convergence: (\sum (-1)^n / n) converges conditionally, not absolutely. -Reality: Test (\sum |a_n|) separately for absolute convergence.
• Fact: Alternating p-series: (\sum (-1)^{n+1} / n^p) converges if (p > 0) (conditionally for (0 < p \leq 1), absolutely for (p > 1)).
• Trap: Misapplying the error bound to non-alternating series: The bound only works for series satisfying AST. -Reality: For other series, use integral/remainder tests or Taylor error bounds.

Rapid-Fire True/False

• Statement: If (\sum (-1)^n a_n) converges, then (\sum a_n) must converge. Answer: FALSE Why the common mistake happens: Confusing conditional convergence (where (\sum a_n) diverges) with absolute convergence.

• Statement: The error bound (|S - S_n| \leq a_{n+1}) guarantees the actual error is always less than (a_{n+1}). Answer: TRUE Why the common mistake happens: Students assume the bound is loose but forget it’s a strict upper limit.

• Statement: The series (\sum (-1)^n (1 + 1/n)) satisfies the Alternating Series Test. Answer: FALSE Why the common mistake happens: Overlooking that (a_n = 1 + 1/n) does not approach 0 (violates (\lim a_n = 0)).