Comparison Test vs. Limit Comparison Test: A Deep Dive into Series Convergence
When determining whether an infinite series converges or diverges, calculus students often encounter two powerful and related tools: the Comparison Test and the Limit Comparison Test. While they share the same fundamental goal, their mechanics, applications, and levels of precision differ significantly. Understanding when and how to use each test is a critical skill for mastering the behavior of infinite series. This article provides a comprehensive breakdown of both tests, illustrating their differences with clear examples and offering a practical guide for choosing the right one.
The Foundation: The Comparison Test
The Comparison Test is the more intuitive of the two. It operates on a simple, logical principle: if you can bound the terms of a series between the terms of two other series whose behavior you already know, you can deduce the behavior of the original series.
There are two primary forms of the Comparison Test:
-
Direct Comparison Test (DCT):
- For Convergence: If you have a series (\sum a_n) with positive terms, and you can find a convergent series (\sum b_n) such that (0 \leq a_n \leq b_n) for all (n) beyond some point, then (\sum a_n) must also converge. The logic is straightforward: if the larger series ((b_n)) has a finite sum, the smaller series ((a_n)), being even smaller, must also have a finite sum.
- For Divergence: Conversely, if you can find a divergent series (\sum c_n) such that (0 \leq c_n \leq a_n) for all (n) beyond some point, then (\sum a_n) must also diverge. If the smaller series ((c_n)) already has an infinite sum, the larger series ((a_n)) must be even larger and therefore also infinite.
Key Takeaway: The DCT requires you to find a suitable "benchmark" series ((b_n) or (c_n)) that is either larger or smaller than your series and whose convergence is already known That alone is useful..
-
Limit Comparison Test (LCT): The LCT is a more refined and often more practical tool. It addresses a limitation of the DCT: what if the terms of your series are not strictly smaller or larger than a known benchmark, but are proportional to it as (n) grows very large?
The LCT states that for two series (\sum a_n) and (\sum b_n) with positive terms, if the limit [ L = \lim_{n \to \infty} \frac{a_n}{b_n} ] exists and is a finite, positive number ((0 < L < \infty)), then either both series converge or both diverge.
This changes depending on context. Keep that in mind.
**Why this works:** The condition \(0 < L < \infty\) means that for sufficiently large \(n\), the terms \(a_n\) and \(b_n\) are essentially proportional to each other. They behave "asymptotically" the same way. If one series' terms are eventually just a constant multiple of the other's, their long-term behavior (convergence or divergence) must be identical.
A Side-by-Side Comparison
To clarify the distinction, let's create a comparative table.
| Feature | Comparison Test (Direct) | Limit Comparison Test |
|---|---|---|
| Core Principle | Inequality-based: (a_n \leq b_n) or (a_n \geq b_n) | Asymptotic Equality: (\lim (a_n / b_n) = L) |
| Requirement | Must find a benchmark series that is strictly larger or smaller. | Must find a benchmark series that is "similar" in behavior. |
| Ease of Application | Can be difficult to find a suitable inequality. Even so, | Often easier, as it only requires finding a similar benchmark. |
| Precision | Less precise; can fail if the inequality is not obvious. Because of that, | More precise; works even if the inequality is not strict. |
| Common Benchmark Series | p-series ((\sum 1/n^p)), geometric series ((\sum ar^n)) | p-series, geometric series, or other known series. |
Illustrative Examples
Let's apply both tests to the same series to see the difference in action.
Example 1: (\sum_{n=1}^{\infty} \frac{1}{n^2 + 5})
-
Using the Direct Comparison Test (DCT): We need to find a benchmark. Notice that for all (n \geq 1), (n^2 + 5 > n^2). So, [ \frac{1}{n^2 + 5} < \frac{1}{n^2} ] The series (\sum \frac{1}{n^2}) is a convergent p-series (p=2 > 1). Since our terms are smaller than the terms of a known convergent series, the DCT tells us that (\sum \frac{1}{n^2 + 5}) converges.
-
Using the Limit Comparison Test (LCT): We choose the benchmark series (\sum b_n = \sum \frac{1}{n^2}), as the (n^2) term dominates for large (n). Now, we compute the limit: [ L = \lim_{n \to \infty} \frac{a_n}{b_n} = \lim_{n \to \infty} \frac{1/(n^2 + 5)}{1/n^2} = \lim_{n \to \infty} \frac{n^2}{n^2 + 5} = \lim_{n \to \infty} \frac{1}{1 + 5/n^2} = 1 ] Since (L = 1), which is finite and positive, and (\sum \frac{1}{n^2}) converges, the LCT confirms that (\sum \frac{1}{n^2 + 5}) also converges.
Example 2: (\sum_{n=1}^{\infty} \frac{n}{n^3 + 2})
-
Using the Direct Comparison Test (DCT): This one is trickier. We might try to say (\frac{n}{n^3 + 2} < \frac{n}{n^3} = \frac{1}{n^2}). This is true, and since (\sum \frac{1}{n^2}) converges, the DCT successfully shows convergence. That said, finding this inequality required a bit of insight.
-
Using the Limit Comparison Test (LCT): For large (n), the series behaves like (\frac{n}{n^3} = \frac{1}{n^2}). So, we choose (b_n = \frac{1}{n^2}). [
