Independent study guide

Series Convergence: Three Decisions to Explain

Explain three series-test decisions, then prove a sequence limit and bound integral-test remainders in a later-course workshop. Seven original questions with hints and explained answers.

Nablio · 7 questionsPublished

Return to Concordia MATH 205 course guide

Practise deciding whether a series converges and explaining why a test applies. These three questions review the ratio test, direct comparison and alternating series. They are for readers who have already met these tests. The aim is to justify convergence or divergence, not to find an exact sum.

In every question, nn starts at 11 and increases through the positive integers. A term ana_n is one item in the series a1+a2+a3+⋯a_1+a_2+a_3+\cdots. A partial sum adds only the first finitely many terms. A series converges when its partial sums approach a finite number as more terms are added.

Try each question before opening its hint or explained answer. First repair a test conclusion, then choose a useful comparison, then combine two convergence checks.

1. When the ratio test is inconclusive

Consider the series with terms

an=1n2+3n.a_n=\frac{1}{n^2+3n}.

A student calculates that an+1/ana_{n+1}/a_n tends to 11 and concludes that the series diverges. What should change in the argument? Give a valid convergence or divergence argument.

Show a starting hint for question 1

First state what the ratio test says when its limit is 11. Then compare the positive terms with 1/n21/n^2. Which direction of inequality would let you prove convergence?

Show the answer and reasoning for question 1
  1. Interpret the ratio-test limit

    The terms are positive, so the absolute-value ratio is the same as an+1/ana_{n+1}/a_n. Substituting the terms and expanding the denominator gives

    n2+3nn2+5n+4.\frac{n^2+3n}{n^2+5n+4}.

    Divide the numerator and denominator by n2n^2:

    1+3/n1+5/n+4/n2.\frac{1+3/n}{1+5/n+4/n^2}.

    As n→∞n\to\infty, the numerator and denominator both approach 1, so the ratio approaches 1.

    When this limit is 11, the ratio test is inconclusive: it proves neither convergence nor divergence. Use another test.

  2. Find a convergent upper comparison

    For every n≥1n\geq1, we have n2+3n≥n2>0n^2+3n\geq n^2>0. Taking reciprocals reverses the inequality, so

    0<1n2+3n≤1n2.0<\frac{1}{n^2+3n}\leq\frac{1}{n^2}.

    A p-series has terms 1/np1/n^p, where pp is a fixed real exponent; it converges when p>1p>1. Here p=2p=2, so the comparison series converges.

  3. State what the comparison proves

    The original terms are nonnegative and no larger than the corresponding terms of a convergent series. Direct comparison therefore proves that the original series converges. Because all its terms are positive, taking absolute values leaves the series unchanged: it also converges absolutely.

    The decisive evidence is the convergent upper comparison, not the inconclusive ratio-test limit.

2. Choose the useful direction of comparison

Consider the series with terms

bn=nn2+4.b_n=\frac{n}{n^2+4}.

A student observes that 0<bn<1/n0<b_n<1/n and says the series must diverge because the harmonic series does. Is that comparison enough? Complete the argument.

Show a starting hint for question 2

The harmonic series is 1+1/2+1/3+⋯1+1/2+1/3+\cdots. To force the partial sums of your series to grow without bound, do you need an upper or a lower bound? Use n≥1n\geq1 to compare n2+4n^2+4 with 5n25n^2.

Show the answer and reasoning for question 2
  1. Explain why the upper bound is insufficient

    The upper bound is correct, but it does not establish divergence. The terms of a convergent series can also be smaller than those of a divergent series: for example, 1/n2≤1/n1/n^2\leq1/n. To prove divergence by direct comparison here, find a lower bound supplied by the terms of a divergent positive series.

  2. Build a divergent lower comparison

    Since n≥1n\geq1, we have 4≤4n24\leq4n^2, so n2+4≤5n2n^2+4\leq5n^2. The denominators are positive; taking reciprocals reverses the inequality. Multiplying by the positive number nn gives

    bn=nn2+4≥15n.b_n=\frac{n}{n^2+4}\geq\frac{1}{5n}.

  3. Compare the partial sums

    The harmonic series, 1+1/2+1/3+⋯1+1/2+1/3+\cdots, has partial sums that grow without bound. The partial sums of the series with terms 1/(5n)1/(5n) are one fifth as large and also grow without bound. The corresponding partial sums of the original series are at least as large.

    The original series therefore diverges to +∞+\infty. A divergent lower comparison forces this conclusion; the student's divergent upper comparison did not.

3. Distinguish convergence from absolute convergence

Consider the series with terms

cn=(−1)n+1n+2.c_n=\frac{(-1)^{n+1}}{n+2}.

Does it converge? Does the series of absolute values converge? Justify both conclusions.

Show a starting hint for question 3

Write the first three terms to see the alternating signs. Then show that ∣cn∣|c_n| decreases for every n≥1n\geq1 and tends to zero as n→∞n\to\infty before using the alternating-series test. Then ask a separate question about the positive terms ∣cn∣=1/(n+2)|c_n|=1/(n+2).

Show the answer and reasoning for question 3
  1. Check the alternating-series conditions

    The series begins 1/3−1/4+1/5−⋯1/3-1/4+1/5-\cdots. Its signs alternate. Its magnitudes are positive and decrease because, for every n≥1n\geq1,

    0<1n+3<1n+2.0<\frac{1}{n+3}<\frac{1}{n+2}.

    They also tend to zero:

    lim⁡n→∞1n+2=0.\lim_{n\to\infty}\frac{1}{n+2}=0.

    These conditions establish convergence by the alternating-series test.

  2. Test the series of absolute values

    Absolute convergence means that the series formed by taking the absolute value of every term converges. Here its terms are ∣cn∣=1/(n+2)|c_n|=1/(n+2). For n≥1n\geq1, we have n+2≤3nn+2\leq3n, and both denominators are positive. Therefore

    1n+2≥13n.\frac{1}{n+2}\geq\frac{1}{3n}.

    The series with terms 1/(3n)1/(3n) diverges because its partial sums are one third of the harmonic partial sums. Direct comparison shows that the series of absolute values also diverges.

  3. Combine the two conclusions

    A series is conditionally convergent when it converges but its series of absolute values diverges. The original alternating series is therefore conditionally convergent.

    Absolute convergence would imply convergence. Here the alternating-series test established convergence, but it did not establish absolute convergence; the comparison above rules that out.

Check your reasoning

Close the answers. For each question, name the test, write the condition or inequality that permits it, and state exactly what it establishes.

  • If question 1 was difficult, review why a ratio-test limit of 11 gives no conclusion.
  • If question 2 was difficult, review the comparison directions for nonnegative terms: an upper bound from a convergent series can prove convergence; a lower bound from a divergent series can prove divergence.
  • If question 3 was difficult, review the alternating-series conditions and the difference between absolute and conditional convergence.

Redo the first question where your reasoning was incomplete, keeping its answer closed. Then choose another series question from your course materials and justify each condition before using a test. The Concordia MATH 205 course guide places these convergence questions alongside the course’s other topics.

Course context and official sources

Concordia's MATH 205 calendar description includes sequences, series and convergence tests. Its Fall 2026 outline lists comparison tests, absolute convergence, the ratio test and alternating series. These sources establish topic relevance; your current section's outline determines the material to study.

These are original Nablio practice questions, not official Concordia exam questions. This focused diagnostic does not cover every convergence test or the whole course. Nablio is independent of Concordia.

Later-course workshop: prove a sequence limit, then bound a series tail

Concordia’s Fall 2026 MATH 205 outline places sequences in week 8 and the integral test in week 9, outside the weeks 1–6 midterm scope. This original workshop is later-course preparation, not early-midterm practice. Your current section’s announced scope remains the guide. You need induction, limits, derivatives and improper integrals.

A sequence (an)(a_n) is a list of terms. A series ∑an\sum a_n concerns the limit of its partial sums. A sequence tending to zero does not by itself make its series converge. For a convergent series with sum S, define SN=∑n=1NanS_N=\sum_{n=1}^N a_n and the remainder RN=S−SN=∑n=N+1∞anR_N=S-S_N=\sum_{n=N+1}^{\infty}a_n. Record the first omitted index before estimating an error.

Worked example 1: justify convergence before solving for a limit

Let a1=1a_1=1 and an+1=2+ana_{n+1}=\sqrt{2+a_n}. Prove convergence and find the limit.

  1. Prove bounds: initially 1≤a1≤21\le a_1\le2. If 1≤an≤21\le a_n\le2, then 1≤2+an≤21\le\sqrt{2+a_n}\le2. Induction gives 1≤an≤21\le a_n\le2 for every n.
  2. Prove monotonicity: for 1≤an≤21\le a_n\le2, both sides of 2+an≥an\sqrt{2+a_n}\ge a_n are nonnegative. Squaring is therefore valid, and the inequality becomes 2+an−an2=(2−an)(an+1)≥02+a_n-a_n^2=(2-a_n)(a_n+1)\ge0. Thus the sequence is nondecreasing.
  3. Establish existence: a nondecreasing sequence bounded above converges. Call its limit L; the bounds imply 1≤L≤21\le L\le2.
  4. Identify it: continuity of the square root gives L=2+LL=\sqrt{2+L}. Squaring yields (L−2)(L+1)=0(L-2)(L+1)=0. The bounds exclude −1, so L = 2.

Solving a fixed-point equation alone identifies possible limits; it does not prove that a sequence converges. The bounds and monotonicity supplied that missing argument.

Worked example 2: turn convergence into a guaranteed error

For S=∑n=1∞1/n2S=\sum_{n=1}^{\infty}1/n^2, find an interval containing S in terms of S_N and choose N that guarantees an absolute error at most 0.01.

The function f(x)=x−2f(x)=x^{-2} is positive, continuous and decreasing for x≥1x\ge1, since f′(x)=−2x−3<0f'(x)=-2x^{-3}<0. Its improper integral converges. Decreasing-function rectangles give

∫N+1∞f(x) dx≤RN≤∫N∞f(x) dx.\int_{N+1}^{\infty}f(x)\,dx\le R_N\le\int_N^{\infty}f(x)\,dx.

The tail starts at N + 1: on [n,n+1], the function lies at or below f(n), while on [n−1,n], it lies at or above f(n). Thus

1N+1≤RN≤1N,SN+1N+1≤S≤SN+1N.\frac1{N+1}\le R_N\le\frac1N,\qquad S_N+\frac1{N+1}\le S\le S_N+\frac1N.

The partial sum underestimates S because omitted terms are positive. N = 100 guarantees ∣S−S100∣≤0.01|S-S_{100}|\le0.01. If a task explicitly requires using this upper bound to establish an error strictly below 0.01, choose N = 101. This is a sufficient guarantee, not a claim about the smallest possible N for the actual error. The term test alone could not have supplied this quantitative bound.

Four reasoning retries

1. A fixed point that the sequence never approaches

Let b1=1b_1=1 and bn+1=3−bnb_{n+1}=3-b_n. Someone solves L = 3 − L and concludes bn→3/2b_n\to3/2. Find the flaw and determine what happens.

Hint

Write the first six terms, then separate odd and even indices.

Explained solution

The terms are 1, 2, 1, 2, 1, 2. Induction gives b2k−1=1b_{2k-1}=1 and b2k=2b_{2k}=2. These subsequences have different limits, so the sequence does not converge. It is bounded, but not monotone. The equation L = 3 − L is only a necessary condition if a limit exists; existence was never proved.

2. Keep the shifted index

Define S=∑n=1∞1/(n+1)2S=\sum_{n=1}^{\infty}1/(n+1)^2, with S_N summing n = 1 through N. Give lower and upper integral bounds for R_N.

Hint

Use f(x)=1/(x+1)2f(x)=1/(x+1)^2, or reindex and identify the last included denominator.

Explained solution

The first omitted term is 1/(N+2)21/(N+2)^2. The integral bounds are 1/(N+2)≤RN≤1/(N+1)1/(N+2)\le R_N\le1/(N+1). Indeed, ∫N+1∞(x+1)−2 dx=1/(N+2)\int_{N+1}^{\infty}(x+1)^{-2}\,dx=1/(N+2) and the integral starting at N is 1/(N+1). Reusing 1/(N+1) as the lower bound would be incorrect: for this strictly decreasing function, the remainder is strictly less than 1/(N+1).

3. Conditions are not a convergence conclusion

A student notes that f(x)=x/(x2+1)f(x)=x/(x^2+1) is positive, continuous and nonincreasing for x ≥ 1, and claims that ∑n=1∞n/(n2+1)\sum_{n=1}^{\infty}n/(n^2+1) therefore converges. What still needs checking?

Hint

Evaluate the improper integral; the integral test can establish divergence too.

Explained solution

f′(x)=(1−x2)/(x2+1)2≤0f'(x)=(1-x^2)/(x^2+1)^2\le0 for x ≥ 1, so the test applies. But ∫1bx/(x2+1) dx=12ln⁡(b2+1)−12ln⁡2→∞\int_1^b x/(x^2+1)\,dx=\tfrac12\ln(b^2+1)-\tfrac12\ln2\to\infty. The series diverges. There is no finite sum S and hence no finite tail-error guarantee of the kind used in worked example 2. Checking monotonicity permits the test; it does not decide its result.

4. Choose N from an error tolerance

For ∑n=1∞n−3\sum_{n=1}^{\infty}n^{-3}, use the integral bound to guarantee RN≤0.001R_N\le0.001. Give the smallest positive integer N meeting this upper-bound criterion.

Hint

Integrate x⁻³ from N to infinity, then solve the resulting inequality before rounding upward.

Explained solution

The positive, continuous function x⁻³ decreases for x ≥ 1, and RN≤∫N∞x−3 dx=1/(2N2)R_N\le\int_N^{\infty}x^{-3}\,dx=1/(2N^2). We need N2≥500N^2\ge500, so N ≥ √500 ≈ 22.36 and the smallest integer satisfying this criterion is 23. Its upper bound is 1/1058 ≈ 0.000945. N = 22 gives 1/968 ≈ 0.001033, so this bound does not certify that choice; that does not by itself prove its actual error exceeds the tolerance.

For each argument, state what is converging, prove every needed condition, then name the first omitted term. Return to comparison reasoning if you need to choose a convergence test, or to the MATH 205 course guide to choose the appropriate study stage. These are original Nablio questions, not official exam material. Tutoring is currently offered in English.

Useful for

Work on the step you’re missing.

Tell us the subject and your preferred time. Nablio confirms the details by text or email.

Request a lesson