Scribe logoScribeNotes & practice
Sign inStart free
  1. Notes
  2. ›AP Calculus BC
  3. ›10 Infinite Sequences and Series
  4. ›10.10 Alternating series error bound

AP Calculus BC · Unit 10 · Infinite Sequences and Series

10.10 Alternating series error bound

Estimate the error in a partial sum of an alternating series and use the next term to set error limits.

  • 4 min read
  • 5 parts
  • 10 questions
  • Draft
  • BC only
Start a Quick 5 on this topic→Jump to the 10 questions
1/5Stopping a series early leaves an error we can estimate
  1. 1Stopping a series early leaves an error we can estimate
  2. 2The error is no larger than the first omitted term
  3. 3The bound holds only when three conditions are met
  4. 4The true sum lies between two consecutive partial sums
  5. 5The same bound tells you how many terms to add
  6. Check your recall
  7. Practice: 10 questions
1

Part 1 of 5: Stopping a series early leaves an error we can estimate

Nobody can add infinitely many terms by hand. In practice you add the first few terms and stop. The result is only an estimate of the full sum.

The sum of the first nnn terms is the partial sum SnS_nSn​. Write the true sum of the whole series as SSS. The error is the true sum minus the estimate, S−SnS-S_nS−Sn​. We usually care about its size, ∣S−Sn∣|S-S_n|∣S−Sn​∣.

Most series have no tidy exact sum. So we need a way to limit the error without knowing SSS.

Exam tip

Count the terms you actually added. S3S_3S3​ means the first three terms, whatever the starting index of the series.

Key terms · tap one for its definition

partial sum

The partial sum SnS_nSn​ is the sum of the first nnn terms of a series, used as an estimate of the whole sum.

error

The error is the true sum minus the partial sum, S−SnS-S_nS−Sn​, and its size is ∣S−Sn∣|S-S_n|∣S−Sn​∣.

Part 1 done

2

Part 2 of 5: The error is no larger than the first omitted term

Now think about what each new term does. When the signs alternate and the sizes shrink, each term pushes the partial sum across the true sum. The pushes get smaller and smaller.

Write the series as ∑n=1∞(−1)n+1an\sum_{n=1}^{\infty}(-1)^{n+1}a_n∑n=1∞​(−1)n+1an​ with every an>0a_n>0an​>0. The first omitted term is the term right after the last one you added. For SnS_nSn​, its size is an+1a_{n+1}an+1​.

The alternating series error bound puts a limit on the error. It says that ∣S−Sn∣≤an+1|S-S_n|\le a_{n+1}∣S−Sn​∣≤an+1​.

In words, the error is at most the size of the first omitted term. You do not need to know SSS to use it.

Key terms · tap one for its definition

first omitted term

The first omitted term is the term an+1a_{n+1}an+1​ that comes immediately after the last term included in SnS_nSn​.

alternating series error bound

The alternating series error bound states that ∣S−Sn∣≤an+1|S-S_n|\le a_{n+1}∣S−Sn​∣≤an+1​ when the terms alternate in sign and their sizes decrease toward zero.

Part 2 done

Quick check

One question on the part you just read

Easy

Identify the term whose size is the alternating series error bound for SnS_nSn​, when the three conditions hold.
Show answer and explanation

Answer: A · Worked explanation

The bound is ∣S−Sn∣≤an+1|S-S_n|\le a_{n+1}∣S−Sn​∣≤an+1​, so it is the size of the first omitted term. Choosing ana_nan​ confuses the last term added with the next one. Choosing Sn+1S_{n+1}Sn+1​ mistakes a sum for a single term, and an average of two terms has no role in the bound at all.
3

Part 3 of 5: The bound holds only when three conditions are met

The bound is not a rule for every series. It depends on three conditions, so check them first. If one fails, the bound may be false.

  • The terms alternate in sign.
  • The sizes ana_nan​ decrease.
  • The sizes ana_nan​ approach zero.

The same conditions make the series converge, so the sum SSS exists. Without convergence, there is no true sum to estimate.

Common mistake

Do not use ana_nan​, the last term added, as the bound. The bound uses an+1a_{n+1}an+1​, the first term left out.

Part 3 done

4

Part 4 of 5: The true sum lies between two consecutive partial sums

Now apply the bound to a concrete series. Take ∑n=1∞(−1)n+1n2=1−14+19−116+⋯\sum_{n=1}^{\infty}\frac{(-1)^{n+1}}{n^2}=1-\frac14+\frac19-\frac1{16}+\cdots∑n=1∞​n2(−1)n+1​=1−41​+91​−161​+⋯. Its terms alternate, and the sizes 1n2\frac1{n^2}n21​ decrease toward zero, so the conditions hold.

Worked example

Bounding the error in a partial sum

Use S3S_3S3​ to estimate ∑n=1∞(−1)n+1n2\sum_{n=1}^{\infty}\frac{(-1)^{n+1}}{n^2}∑n=1∞​n2(−1)n+1​. Find the error bound and an interval that contains the sum.
  1. 1
    Add the first three terms: S3=1−14+19=36−9+436=3136≈0.8611S_3=1-\frac14+\frac19=\frac{36-9+4}{36}=\frac{31}{36}\approx0.8611S3​=1−41​+91​=3636−9+4​=3631​≈0.8611.
  2. 2
    The first omitted term is the fourth, with size a4=142=116=0.0625a_4=\frac{1}{4^2}=\frac1{16}=0.0625a4​=421​=161​=0.0625.
  3. 3
    So ∣S−S3∣≤0.0625|S-S_3|\le0.0625∣S−S3​∣≤0.0625.
  4. 4
    Subtract and add the bound: 0.8611−0.0625=0.79860.8611-0.0625=0.79860.8611−0.0625=0.7986 and 0.8611+0.0625=0.92360.8611+0.0625=0.92360.8611+0.0625=0.9236.

Answer: The error is at most 116\frac1{16}161​, so 0.799≤S≤0.9240.799\le S\le0.9240.799≤S≤0.924.

The bound also gives a sharper result. Consecutive partial sums lie on opposite sides of SSS. So the sum is between SnS_nSn​ and Sn+1S_{n+1}Sn+1​.

Here S4=115144≈0.799S_4=\frac{115}{144}\approx0.799S4​=144115​≈0.799, so 0.799≤S≤0.8610.799\le S\le0.8610.799≤S≤0.861. This interval has width 0.0625, half the width of the first one. The first omitted term, −116-\frac1{16}−161​, is negative. That makes S3S_3S3​ an overestimate, a value above the true sum. A positive first omitted term makes SnS_nSn​ an underestimate.

Figure 1. The partial sums alternate above and below the true sum, and each swing is smaller than the one before.

Part 4 done

Quick check

One question on the part you just read

Easy

Determine the alternating series error bound when S4S_4S4​ estimates ∑n=1∞(−1)n+1n2\sum_{n=1}^{\infty}\frac{(-1)^{n+1}}{n^2}∑n=1∞​n2(−1)n+1​.
Show answer and explanation

Answer: B · Worked explanation

The first omitted term after S4S_4S4​ is the fifth, with size a5=152=125=0.04a_5=\frac{1}{5^2}=\frac1{25}=0.04a5​=521​=251​=0.04. So ∣S−S4∣≤0.04|S-S_4|\le0.04∣S−S4​∣≤0.04, which is the answer. The value 116\frac1{16}161​ is a4a_4a4​, the size of the last term added. That is the bound for S3S_3S3​ rather than for S4S_4S4​, so it is the most tempting wrong value.
5

Part 5 of 5: The same bound tells you how many terms to add

Often a problem fixes the accuracy you need. Then you work backward from the bound. Require an+1a_{n+1}an+1​ to be at most the allowed error, then solve for nnn.

Worked example

Finding the number of terms

How many terms of ∑n=1∞(−1)n+1n2\sum_{n=1}^{\infty}\frac{(-1)^{n+1}}{n^2}∑n=1∞​n2(−1)n+1​ guarantee an error of at most 0.001?
  1. 1
    We need an+1=1(n+1)2≤0.001a_{n+1}=\frac{1}{(n+1)^2}\le0.001an+1​=(n+1)21​≤0.001.
  2. 2
    Rearrange: (n+1)2≥1000(n+1)^2\ge1000(n+1)2≥1000.
  3. 3
    Since 312=96131^2=961312=961 and 322=102432^2=1024322=1024, we need n+1≥32n+1\ge32n+1≥32.
  4. 4
    So n≥31n\ge31n≥31.

Answer: Adding 31 terms guarantees an error of at most 0.001.

Notice how the answer scales. The bound shrinks like 1n2\frac1{n^2}n21​ for this series. So cutting the allowed error by a factor of 10 needs only about three times as many terms.

The bound is a guarantee, not a prediction. The actual error may be smaller, so fewer terms might do. The bound alone cannot promise that.

Exam tip

When you justify a bound, name the conditions you checked. Then write ∣S−Sn∣≤an+1|S-S_n|\le a_{n+1}∣S−Sn​∣≤an+1​ and give its value.

Go deeperWhy the bound holds

Look at the leftover tail with its sign removed, T=an+1−an+2+an+3−⋯T=a_{n+1}-a_{n+2}+a_{n+3}-\cdotsT=an+1​−an+2​+an+3​−⋯. Its size is the size of the error. Group it as (an+1−an+2)+(an+3−an+4)+⋯(a_{n+1}-a_{n+2})+(a_{n+3}-a_{n+4})+\cdots(an+1​−an+2​)+(an+3​−an+4​)+⋯. Each bracket is at least 0 because the sizes decrease, so T≥0T\ge0T≥0.

Now group it as an+1−(an+2−an+3)−(an+4−an+5)−⋯a_{n+1}-(a_{n+2}-a_{n+3})-(a_{n+4}-a_{n+5})-\cdotsan+1​−(an+2​−an+3​)−(an+4​−an+5​)−⋯. Each subtracted bracket is at least 0, so T≤an+1T\le a_{n+1}T≤an+1​. Together, 0≤T≤an+10\le T\le a_{n+1}0≤T≤an+1​, which is the bound.

Note read

Check your recall[13 marks]

6 written questions. Answer each one in your head or on paper, then open the mark scheme.

  1. 1.Describe[2 marks]

    Describe what the alternating series error bound says about ∣S−Sn∣|S-S_n|∣S−Sn​∣ for a series that meets its three conditions.

    Show mark schemeHide mark scheme

    Model answer

    The size of the error is at most the size of the first omitted term, so ∣S−Sn∣≤an+1|S-S_n|\le a_{n+1}∣S−Sn​∣≤an+1​.

    • 1 mark for stating that the size of the error is at most the size of a particular term
    • 1 mark for identifying that term as the first omitted term, an+1a_{n+1}an+1​
  2. 2.Determine[2 marks]

    Determine the error bound when S4S_4S4​ is used to estimate ∑n=1∞(−1)n+1n3\sum_{n=1}^{\infty}\frac{(-1)^{n+1}}{n^3}∑n=1∞​n3(−1)n+1​.

    Show mark schemeHide mark scheme

    Model answer

    The first omitted term is the fifth, so a5=153=1125=0.008a_5=\frac{1}{5^3}=\frac1{125}=0.008a5​=531​=1251​=0.008, and ∣S−S4∣≤0.008|S-S_4|\le0.008∣S−S4​∣≤0.008.

    • 1 mark for identifying the first omitted term as a5=153a_5=\frac{1}{5^3}a5​=531​
    • 1 mark for giving the bound as 1125=0.008\frac1{125}=0.0081251​=0.008
  3. 3.Justify[3 marks]

    Justify that the alternating series error bound applies to ∑n=1∞(−1)n+11n\sum_{n=1}^{\infty}(-1)^{n+1}\frac{1}{\sqrt n}∑n=1∞​(−1)n+1n​1​.

    Show mark schemeHide mark scheme

    Model answer

    The factor (−1)n+1(-1)^{n+1}(−1)n+1 makes the terms alternate in sign. The sizes 1n\frac1{\sqrt n}n​1​ decrease as nnn increases. They also approach 0, so all three conditions hold.

    • 1 mark for showing that the terms alternate in sign
    • 1 mark for showing that the sizes 1n\frac1{\sqrt n}n​1​ decrease
    • 1 mark for showing that the sizes approach 0
  4. 4.Find[2 marks]

    Find the smallest number of terms that guarantees an error of at most 0.02 for ∑n=1∞(−1)n+1n\sum_{n=1}^{\infty}\frac{(-1)^{n+1}}{n}∑n=1∞​n(−1)n+1​.

    Show mark schemeHide mark scheme

    Model answer

    We need 1n+1≤0.02\frac1{n+1}\le0.02n+11​≤0.02, so n+1≥50n+1\ge50n+1≥50 and n≥49n\ge49n≥49. The smallest number of terms is 49.

    • 1 mark for setting an+1=1n+1≤0.02a_{n+1}=\frac1{n+1}\le0.02an+1​=n+11​≤0.02
    • 1 mark for solving to get 49 terms
  5. 5.Predict[2 marks]

    Predict whether S4S_4S4​ overestimates or underestimates ∑n=1∞(−1)n+1n2\sum_{n=1}^{\infty}\frac{(-1)^{n+1}}{n^2}∑n=1∞​n2(−1)n+1​, and give a reason.

    Show mark schemeHide mark scheme

    Model answer

    S4S_4S4​ underestimates the sum. The first omitted term, +125+\frac1{25}+251​, is positive, so the true sum is above S4S_4S4​.

    • 1 mark for predicting that S4S_4S4​ is an underestimate
    • 1 mark for using the positive sign of the first omitted term as the reason
  6. 6.Explain[2 marks]

    Explain why the true sum lies between SnS_nSn​ and Sn+1S_{n+1}Sn+1​ for such a series.

    Show mark schemeHide mark scheme

    Model answer

    Each added term pushes the partial sum across the true sum by a smaller amount. So consecutive partial sums lie on opposite sides of the sum.

    • 1 mark for stating that consecutive partial sums lie on opposite sides of the true sum
    • 1 mark for linking this to alternating signs and shrinking terms

Practice this topic

10 quick-fire multiple-choice questions, easy to hard. Tap an answer to see if you're right and the worked explanation. (The written questions with mark schemes are under “Check your recall” above.)

0 right0 of 10 answered

Includes the 2 quick checks in the note above.

3.

Easy·Identify[1 mark]

Identify the condition that is NOT needed for the alternating series error bound.
Show answer and explanation

Answer: A · Worked explanation

The conditions are alternating signs, decreasing sizes, and sizes that approach 0. The first term may be any size, such as 5, as long as later sizes shrink toward zero. Requiring every term to be below 1 is not part of the conditions. The other three statements each describe one of the three conditions, so they are needed.
4.

Medium·Determine[2 marks]

The first 4 terms of 1−13+15−17+⋯1-\frac13+\frac15-\frac17+\cdots1−31​+51​−71​+⋯ are added. Determine the alternating series error bound.
Show answer and explanation

Answer: B · Worked explanation

The sizes are 12n−1\frac{1}{2n-1}2n−11​. The fourth term has size 17\frac1771​, so the first omitted term is the fifth, with size 19\frac1991​. The error is at most 19\frac1991​. Choosing 17\frac1771​ gives the size of the last term added, which mixes up ana_nan​ with an+1a_{n+1}an+1​.
5.

Medium·Predict[2 marks]

For ∑n=1∞(−1)n+1n2\sum_{n=1}^{\infty}\frac{(-1)^{n+1}}{n^2}∑n=1∞​n2(−1)n+1​, a student uses S5S_5S5​. Predict how S5S_5S5​ compares with the true sum SSS.
Show answer and explanation

Answer: C · Worked explanation

The first omitted term is the sixth, −136-\frac1{36}−361​, which is negative. The error S−S5S-S_5S−S5​ has that sign, so SSS is below S5S_5S5​ and S5S_5S5​ is an overestimate. The last term added is positive, but the omitted term sets the sign. The bound gives a size, and the sign comes from the omitted term.
6.

Medium·Determine[2 marks]

Determine the smallest number of terms nnn that guarantees ∣S−Sn∣≤0.01|S-S_n|\le0.01∣S−Sn​∣≤0.01 for ∑n=1∞(−1)n+1n2\sum_{n=1}^{\infty}\frac{(-1)^{n+1}}{n^2}∑n=1∞​n2(−1)n+1​.
Show answer and explanation

Answer: D · Worked explanation

Require an+1=1(n+1)2≤0.01a_{n+1}=\frac1{(n+1)^2}\le0.01an+1​=(n+1)21​≤0.01, so (n+1)2≥100(n+1)^2\ge100(n+1)2≥100 and n+1≥10n+1\ge10n+1≥10. That gives n≥9n\ge9n≥9, and the omitted term 1100=0.01\frac1{100}=0.011001​=0.01 just meets the limit. For n=8n=8n=8 the omitted term is 181≈0.0123\frac1{81}\approx0.0123811​≈0.0123, which is too large. The value 99 comes from solving 1n+1≤0.01\frac1{n+1}\le0.01n+11​≤0.01 and forgetting the square.
7.

Medium·Interpret[2 marks]

Interpret an alternating series error bound of 0.003 for S6S_6S6​. Which statement is guaranteed?
Show answer and explanation

Answer: D · Worked explanation

A bound gives a limit, not the exact error, so ∣S−S6∣≤0.003|S-S_6|\le0.003∣S−S6​∣≤0.003 is guaranteed. The bound equals the size of the seventh term, not the sixth, so the sixth term is not the number 0.003. It also applies only to S6S_6S6​, because other partial sums have their own bounds and the true sum need not be close to all of them.
8.

Hard·Identify[2 marks]

Identify the series for which the alternating series error bound can be used.
Show answer and explanation

Answer: C · Worked explanation

For the series with nn2+1\frac{n}{n^2+1}n2+1n​, the terms alternate and the sizes 0.5,0.4,0.3,…0.5, 0.4, 0.3, \dots0.5,0.4,0.3,… decrease toward 0. The series with n+1n\frac{n+1}{n}nn+1​ has sizes approaching 1, not 0. The series with 1n2\frac1{n^2}n21​ never changes sign, and the series with (−1)nn(-1)^n n(−1)nn has growing sizes.
9.

Hard·Determine[2 marks]

Determine the smallest number of terms nnn that guarantees ∣S−Sn∣≤0.05|S-S_n|\le0.05∣S−Sn​∣≤0.05 for ∑n=1∞(−1)n+1n\sum_{n=1}^{\infty}\frac{(-1)^{n+1}}{\sqrt n}∑n=1∞​n​(−1)n+1​.
Show answer and explanation

Answer: D · Worked explanation

Require 1n+1≤0.05\frac1{\sqrt{n+1}}\le0.05n+1​1​≤0.05, so n+1≥20\sqrt{n+1}\ge20n+1​≥20 and n+1≥400n+1\ge400n+1≥400. That gives n≥399n\ge399n≥399, since the omitted term 1400=0.05\frac1{\sqrt{400}}=0.05400​1​=0.05 just meets the limit. The value 400 would also work but is not the smallest, because it treats ana_nan​ as the bound. The value 19 comes from forgetting to square the 20.
10.

Hard·Determine[2 marks]

A series meets the three conditions. Its partial sums are S4=0.80S_4=0.80S4​=0.80 and S5=0.76S_5=0.76S5​=0.76. Determine the narrowest interval guaranteed to contain SSS.
Show answer and explanation

Answer: C · Worked explanation

Consecutive partial sums lie on opposite sides of the sum, so SSS is between S5=0.76S_5=0.76S5​=0.76 and S4=0.80S_4=0.80S4​=0.80. The error bound alone gives S4±0.04S_4\pm0.04S4​±0.04, the interval from 0.76 to 0.84. That interval and the other wider ones are true, but they are not the narrowest guarantee.

Keep going

Next topic10.11 Taylor polynomial approximationsPrevious topic10.9 Absolute and conditional convergence

More in Unit 10 · Infinite Sequences and Series

  1. 10.6Comparison testsNotes · 10 questions· BC only
  2. 10.7Alternating series testNotes · 10 questions· BC only
  3. 10.8Ratio testNotes · 10 questions· BC only
  4. 10.9Absolute and conditional convergenceNotes · 10 questions· BC only
  5. 10.10Alternating series error boundNotes · 10 questions· BC only
  6. 10.11Taylor polynomial approximationsNotes · 10 questions· BC only
  7. 10.12Lagrange error boundNotes · 10 questions· BC only
  8. 10.13Radius and interval of convergenceNotes · 10 questions· BC only
  9. 10.14Taylor and Maclaurin seriesNotes · 10 questions· BC only

All AP Calculus BC topics →

Draft: written by Scribe and checked automatically, not yet by a subject reviewer.

Practice questions are original, written by Scribe and checked automatically for correctness.

Topic codes follow the official numbering. Not endorsed by College Board or any exam board.

On this page

0/5 read

  1. 1Stopping a series early leaves an error we can estimate
  2. 2The error is no larger than the first omitted term
  3. 3The bound holds only when three conditions are met
  4. 4The true sum lies between two consecutive partial sums
  5. 5The same bound tells you how many terms to add
  6. Check your recall
  7. Practice: 10 questions
All coursesScribeHow it worksPricingPrivacyTerms

Notes and questions are written by Scribe in our own words. Topic codes follow the official course numbering so you can find your place. Not produced or endorsed by the IB, College Board, AQA, OCR, Pearson Edexcel or any exam board.

© 2026 Scribe

Partial sums close in on the true sum

Partial sum Sₙ012345670.60.70.80.911.1Number of terms n
  • True sum (about 0.822)
  • Partial sums Sₙ