📝 p-series convergence test (40 MCQs)

📖 Explanation: This question targets a fundamental misconception about the boundary behavior of p-series. While it is true that $\lim_{k \to \infty} \frac{1}{k^{0.99}} = 0$, the Divergence Test only provides a necessary condition, not a sufficient one. For p-series specifically, the threshold for convergence is strictly $p > 1$. Since $0.99 < 1$, the series behaves similarly to the harmonic series and diverges, regardless of how close the exponent is to unity. Understanding this sharp boundary is crucial for higher-order analysis.

📖 Explanation: This application question connects the abstract mathematical definition of p-series to a physical modeling scenario. The series given is precisely the Riemann zeta function definition for real inputs. In physics, finite total energy requires the sum to converge. Based on the convergence criteria for p-series, the sum is finite if and only if the exponent $x$ is strictly greater than 1. At $x=1$, the harmonic series diverges logarithmically, implying infinite energy, so the domain must be restricted to $x > 1$.

📖 Explanation: This error analysis question addresses a very common mistake when applying the Integral Test. Students often conflate the *test for convergence* with a *method for finding the sum*. The Integral Test states that the series and the improper integral share the same convergence behavior (both converge or both diverge). However, their values are generally distinct. In this case, while the integral evaluates to 1, the actual sum of the series is $\pi^2/6 \approx 1.645$. Recognizing this distinction is vital for accurate mathematical reasoning.

📖 Explanation: This mixed concept question requires comparing a standard p-series with a logarithmic variant. Series A is a p-series with $p = 1.001 > 1$, so it converges. Series B is not a pure p-series but can be analyzed via the Integral Test; the substitution $u = \ln k$ transforms it into a convergent p-integral. This highlights a nuanced concept: while $1/k$ diverges, adding a squared logarithmic denominator forces convergence. Both converge, challenging the intuition that only $p>1$ polynomial terms guarantee summability.

📖 Explanation: This graph-based interpretation question tests the ability to visualize divergence rates. If partial sums grow without bound, the series diverges, eliminating $p > 1$. If $p \leq 0$, terms do not approach zero, causing linear or explosive growth rather than slowing progressive growth. The description 'slows down progressively but never plateaus' characterizes the sub-linear divergence of p-series where $0 < p \leq 1$. Specifically, $p=1$ grows logarithmically, and $0 < p < 1$ grows as $n^{1-p}$. Thus, the parameter must lie in $(0, 1]$.

📖 Explanation: This application/scenario question links theoretical convergence speed to practical computation. Convergence rate for p-series depends heavily on how much $p$ exceeds 1. Larger $p$ values yield rapidly decaying terms and fast convergence. As $p$ approaches 1 from above, the tail of the series decays extremely slowly, requiring millions of terms for modest accuracy. Therefore, $p = 1.01$ represents the most computationally expensive case among the options, illustrating the practical consequences of the theoretical boundary at $p=1$.

📖 Explanation: This conceptual understanding question tests knowledge of the precise boundary conditions for p-series. The harmonic series ($p=1$) is the critical threshold. Option D fails because removing finite terms doesn't affect convergence. Option C yields the same divergent series. Option A creates a conditionally convergent series only under specific interpretations, but strictly speaking, modifying the exponent is the definitive way to force absolute convergence. Any increase in the exponent, no matter how small ($\epsilon > 0$), pushes the series into the convergent regime $p > 1$.

📖 Explanation: This multi-step reasoning question validates the correct application of the Limit Comparison Test within the context of p-series. The student correctly identified a suitable p-series benchmark ($p=2$). Finding a finite, positive limit $\rho = 1$ confirms that both series share the same convergence behavior. Since the benchmark p-series converges ($2 > 1$), the original series must also converge. This reinforces that LCT relies on asymptotic equivalence, not exact equality, and validates using p-series as standard comparators.

📖 Explanation: This direct recall/conceptual question checks understanding of test prerequisites. The p-series test applies strictly to series of the form $\sum 1/k^p$. The given series has $\sqrt{k^2+1}$ in the denominator, which is asymptotically equivalent to $k$ but algebraically distinct. One must first use the Limit Comparison Test with $1/k$ to establish behavioral equivalence before invoking p-series properties. Recognizing when a test is *not* directly applicable prevents mechanical misapplication and encourages proper analytical sequencing.

📖 Explanation: This challenging/mixed concept question extends p-series logic to logarithmic scales via the Integral Test. Substituting $u = \ln k$ transforms the integral into $\int \frac{du}{u^p}$, which is now a standard p-integral. This transformed integral converges only when $p > 1$. This result is counterintuitive for students who assume logarithms always improve convergence; here, the log acts as the primary variable, making $p=1$ the new divergence threshold. It demonstrates how structural transformations reveal hidden p-series behaviors.

📖 Explanation: This multi-concept question explores closure properties of convergent p-series. If $\sum a_k$ converges as a p-series, then $a_k = 1/k^p$ with $p > 1$. Option A is an alternating version of an absolutely convergent series, so it converges. Option B becomes $\sum 1/k^{2p}$; since $p > 1$, $2p > 2 > 1$, ensuring convergence. Option C becomes $\sum 1/k^{p/2}$, which may diverge if $1 < p \leq 2$. Thus, both A and B are guaranteed to converge, testing understanding of exponent manipulation.

📖 Explanation: This application/reasoning question emphasizes strategic problem-solving over mechanical testing. For rational functions, the Ratio Test is typically inconclusive (limit = 1). Partial fractions are unnecessarily complex. The most efficient HOTS approach is asymptotic analysis: numerator ~ $n^2$, denominator ~ $n^4$, ratio ~ $1/n^2$. This immediately suggests the Limit Comparison Test with the convergent p-series $p=2$. Developing this intuition for dominant balance is more valuable than memorizing test algorithms.

📖 Explanation: This conceptual comparison question probes deeper understanding of divergence *rates*. Many students treat divergence as binary, but the rate matters for applications. The harmonic series diverges logarithmically ($\ln n$), extremely slowly. The series with $p=0.5$ diverges as $n^{0.5}$, which is polynomial and significantly faster. Recognizing this qualitative difference helps in error estimation and modeling. Option B captures this nuance, distinguishing between types of divergence beyond simple pass/fail convergence tests.

📖 Explanation: This error analysis question addresses misconceptions about subseries. A common fallacy is thinking that removing terms could turn a convergent series into a divergent one. In reality, if a series of positive terms converges, *any* subseries (including odd-only terms) must also converge by the Comparison Test. Since $\sum 1/k^3$ converges ($p=3 > 1$), the odd-term subset definitely converges. The student's reasoning fundamentally misunderstands the monotonicity of partial sums for positive series.

📖 Explanation: This Olympiad-style estimation question requires connecting numerical magnitude to theoretical parameters. For $p=2$, the sum approaches $\pi^2/6 \approx 1.645$, far below 5. For $p=0.5$, $S_{1000} \approx 2\sqrt{1000} \approx 63$, far above 5. For $p=1$, $S_{1000} \approx \ln(1000) + \gamma \approx 7.5$. A value of 5.0 with slow growth suggests $p$ is just above 1, where convergence is extremely gradual and partial sums remain large even at high n. This inverse reasoning from data to parameter is advanced analytical thinking.

📖 Explanation: This conceptual foundation question ensures understanding of domain restrictions. If $p < 0$, let $p = -q$ where $q > 0$; terms become $k^q$, which grow without bound, failing the Divergence Test. If $p = 0$, every term is 1, clearly diverging. Thus, meaningful convergence discussion only occurs for $p > 0$. Understanding why definitions include constraints prevents blind formula application and builds rigorous mathematical habits.

📖 Explanation: This scenario-based modeling question highlights sensitivity analysis near critical thresholds. At nominal $\delta = 0$, the series is exactly harmonic and diverges. With uncertainty $\pm 0.001$, the true exponent could be $0.999$ (divergent) or $1.001$ (convergent). This bistability at the boundary means the model's fundamental validity (finite vs infinite energy) is uncertain. Engineers must recognize that operating at critical exponents with measurement error creates existential model risk, not just quantitative imprecision.

📖 Explanation: This application question tests proper setup of the Integral Test. The integrand must exactly match the series term with $k$ replaced by $x$. Options A and B represent different series. Option D represents a different logarithmic power. Only Option C correctly mirrors $\frac{1}{k^2 \ln k}$. Furthermore, this integral converges because $1/x^2$ dominates $1/\ln x$, making it comparable to $p=2$. Correct formulation is the essential first step before evaluation.

📖 Explanation: This error analysis question targets terminology precision. The student correctly determined convergence via p-test on absolute values. However, when $\sum |a_k|$ converges, the original series converges *absolutely*, not conditionally. Conditional convergence specifically means the series converges but $\sum |a_k|$ diverges. Confusing these terms indicates incomplete understanding of convergence classification hierarchy. Absolute convergence is stronger and implies unconditional rearrangement safety, unlike conditional convergence.

📖 Explanation: This graph-based comparison reinforces visual intuition for convergence rates. Both series converge since $p > 1$, so partial sums are increasing and bounded above. The key difference is rate: $p=3$ terms decay as $1/k^3$, reaching near-limit within dozens of terms. $p=1.1$ decays as $1/k^{1.1}$, requiring thousands of terms for similar proximity. Graphically, the $p=3$ curve flattens quickly while $p=1.1$ maintains noticeable slope longer. Visualizing this helps calibrate expectations for numerical approximation.

📖 Explanation: This method-selection question tests procedural knowledge. The Alternating Series Estimation Theorem *only* applies to alternating series. Since $\sum 1/k^4$ has all positive terms, that theorem is inapplicable regardless of bound tightness. The Integral Test Remainder Estimate is specifically designed for positive, decreasing functions like p-series. Recognizing test applicability conditions prevents wasted effort and incorrect error bounds. This basic discrimination is foundational for effective series analysis.

📖 Explanation: This conceptual boundary question tests precise knowledge of the convergence/divergence threshold. P-series diverge for $p \leq 1$ and converge for $p > 1$. Therefore, among divergent p-series, the supremum of possible p values is exactly 1. Values like 1.0001 would converge. Understanding that $p=1$ is the *maximum* divergent exponent (not just a boundary point) clarifies the topology of the parameter space and prevents off-by-one errors in classification.

📖 Explanation: This Olympiad-level diagnostic question bridges theory and computation. Mathematically, $p = 1.0000001 > 1$ guarantees convergence. However, CAS might parse the decimal as rational approximation or suffer precision loss, treating it as harmonic. Alternatively, the partial sum grows so slowly that for $N=10^9$, $S_N \approx \ln(N)/10^{-7}$ could appear unbounded numerically. But the *most likely* systemic issue is parsing/exactness. This highlights that computational outputs require theoretical verification; blind trust in software can lead to false conclusions about borderline cases.

📖 Explanation: This multi-step reasoning question explores flexibility in analysis. Factoring reveals the $1/k^2$ structure explicitly. Substitution reindexes to a shifted p-series. Direct LCT with $1/k^2$ bypasses algebraic manipulation entirely. All three approaches legitimately connect the given series to p-series theory. Recognizing multiple valid pathways enhances problem-solving adaptability. No single method is universally superior; context determines efficiency. This meta-cognitive awareness is characteristic of advanced mathematical thinking.

📖 Explanation: This conceptual understanding question reinforces scale invariance. Theorem: $\sum c \cdot a_k$ converges iff $\sum a_k$ converges for $c \neq 0$. Thus, $\sum 5/k^2$ and $\sum 1/k^2$ share identical convergence status. Constants scale the limit value but cannot convert divergence to convergence or vice versa. This principle simplifies analysis by allowing focus on asymptotic form rather than coefficients. Internalizing this prevents unnecessary algebraic cleanup before applying tests.

📖 Explanation: This conceptual clarification addresses hidden assumptions. The p-series definition $\sum 1/k^p$ accepts any real $p$. Convergence criterion $p > 1$ holds for all reals, rational or irrational. Since $\pi \approx 3.14 > 1$, the series converges absolutely. Some students mistakenly believe calculus operations require rational exponents, but series convergence is a topological property independent of number-theoretic classification. Clarifying this expands conceptual scope beyond textbook examples.

📖 Explanation: This error analysis question targets the most frequent Comparison Test mistake. The test states: if smaller diverges → larger diverges; if larger converges → smaller converges. Knowing that a *larger* series diverges tells us nothing about a smaller series (it could converge or diverge). Here, $1/k^2 < 1/k$ and larger diverges gives no information. The student applied the test backwards. Mastering directional logic is essential for valid comparative reasoning.

📖 Explanation: This application question develops asymptotic modeling skills. For large $k$, $k^2 + k \sim k^2$, so terms behave as $1/k^2$. This suggests $\sum 1/k^2$ as the appropriate approximant. Note that while $\sum 1/(k(k+1))$ telescopes exactly to 1, the question asks for *asymptotic approximation* via p-series form. Choosing $1/k^2$ reflects understanding of dominant balance. This skill transfers to perturbation theory and numerical analysis where exact solutions are unavailable.

📖 Explanation: This Olympiad-style inverse problem connects sum values to parameters. We know $\zeta(2) = \pi^2/6 \approx 1.645$ and $\zeta(1^+) \to \infty$. Since $\zeta(p)$ is strictly decreasing for $p > 1$, and $2.5 > 1.645$, the corresponding p must be less than 2. Also, since the sum is finite, $p > 1$. Thus $1 < p < 2$. This requires knowing specific zeta values and monotonicity properties, representing sophisticated synthesis of analytic number theory concepts within calculus framework.

📖 Explanation: This mixed concept question tests algebraic closure properties. Adding a convergent p-series ($p=2$) to a divergent harmonic series yields divergence, since convergent + divergent = divergent. This shows convergence is not preserved under arbitrary addition. Options B and C preserve convergence. Option D involves conditional convergence complications. Understanding non-closure prevents erroneous assumptions like 'adding convergent series always yields convergent series.' This algebraic awareness is crucial for manipulating infinite expressions safely.

📖 Explanation: This technical detail question ensures rigorous application. At $x=1$, $\ln x = 0$, making the integrand undefined. Starting at 2 avoids this singularity. Additionally, many textbooks define this series starting at 2 for precisely this reason. The Integral Test also requires continuity and positivity on $[a, \infty)$, violated at $x=1$. All listed reasons are valid and interconnected. Attention to domain restrictions prevents invalid integral evaluations and demonstrates mathematical care.

📖 Explanation: This conceptual correction addresses intuitive but wrong heuristics. Students often associate roots with 'smallness,' but $1/\sqrt{k}$ decays slower than $1/k$. Expressing $\sqrt{k}$ as $k^{0.5}$ reveals $p = 0.5$. Since $0.5 < 1$, the p-series test definitively establishes divergence. Translating radical notation to exponential form is a fundamental skill that unlocks systematic analysis. Correcting this misconception prevents systematic errors across many problems involving fractional exponents.

📖 Explanation: This cross-topic connection links integration and series. The Integral Test proof literally constructs upper/lower Riemann sums bounding the integral using series terms. Partial sums $S_n = \sum_{k=1}^n k^{-p}$ approximate the area under $x^{-p}$. This geometric interpretation explains why series and integral share convergence: both measure the same underlying area. Understanding this duality enriches comprehension beyond symbolic manipulation and provides visual intuition for abstract convergence criteria.

📖 Explanation: This true/false analysis tests depth of knowledge. Options A, B, and D are standard facts. Option C is false: $\zeta(2) = \pi^2/6$ is irrational, as are most zeta values. Only specific even integers yield rational multiples of powers of $\pi$, never rationals themselves (except trivially). Believing sums are rational reflects limited exposure to transcendental number theory. Identifying this falsehood requires distinguishing between convergence behavior and arithmetic properties of limits.

📖 Explanation: This application question practices remainder estimation. Computing $\int_n^\infty x^{-3/2} dx = [-2x^{-1/2}]_n^\infty = 2/\sqrt{n}$. This matches option C. Note that options B and D are algebraically equivalent to C, but C is the standard simplified form. Option A misses the coefficient. Proper remainder bounds enable error-controlled numerical approximation. Deriving this bound reinforces the connection between integral calculus and series error analysis, a key practical skill in scientific computing.

📖 Explanation: This conceptual limitation question clarifies test scope. The p-series test applies exclusively to series with positive terms of form $1/k^p$. The presence of $\sin k$ introduces sign changes and non-monotonicity. While the series *does* converge (by Absolute Convergence Test since $|\sin k / k^2| \leq 1/k^2$), this conclusion requires additional machinery beyond the basic p-test. Recognizing when a test's hypotheses fail prevents misapplication and guides selection of appropriate alternatives like comparison on absolute values.

📖 Explanation: This Olympiad-level asymptotic question quantifies convergence rate. The asymptotic formula shows remainder decays as power law $n^{1-p}$. As p increases, exponent $1-p$ becomes more negative, accelerating decay. For $p=2$, remainder ~ $1/n$; for $p=3$, ~ $1/n^2$. This power-law dependence contrasts sharply with exponential decay in geometric series. Understanding this scaling enables informed algorithm design and error budgeting in numerical methods involving p-series tails.

📖 Explanation: This scenario-based decision question applies theory to risk management. With $p = 1.0 \pm 0.1$, true p could be 0.9 (divergent/infinite risk). Regulatory compliance requires guaranteed finiteness, not probabilistic likelihood. Operating at critical boundary with uncertainty is unacceptable. Recommendation must be rejection or redesign until $p > 1$ is confirmed with margin. This illustrates how mathematical boundaries translate to engineering safety margins and why 'approximately convergent' is meaningless in compliance contexts.

📖 Explanation: This nuanced concept question explores LCT edge cases. When $\rho = \lim (a_k/b_k) = 0$, $a_k$ is negligible relative to $b_k$. If $b_k$ converges, $a_k$ must converge. But if $b_k$ diverges, $a_k$ could converge or diverge. Pair B: $(1/k^3)/(1/k^2) = 1/k \to 0$; both converge. Pair A: $(1/k^2)/(1/k) = 1/k \to 0$; numerator converges, denominator diverges. Thus pair A shows ρ=0 permits different outcomes when comparator diverges. Understanding this asymmetry prevents overgeneralization of LCT.

📖 Explanation: This mixed concept question combines power series and p-series. Ratio test gives R=1. At x=1, series becomes $\sum 1/k^2$, convergent p-series (p=2). At x=-1, series becomes $\sum (-1)^k/k^2$, absolutely convergent since $\sum 1/k^2$ converges. Thus convergence at both endpoints. This synthesis requires recognizing that endpoint evaluation reduces to p-series analysis. Many students check R correctly but mishandle endpoints by forgetting absolute convergence implications. This integrated skill is essential for complete power series characterization.