Leibniz Test Made Easy: Alternating Series Convergence Explained
The Leibniz test, often called the alternating series test, is one of the first convergence tools students meet in calculus. It applies to series whose terms alternate in sign and whose positive magnitudes steadily shrink toward zero. Named after Gottfried Wilhelm Leibniz, the test offers a remarkably simple way to prove that many alternating series converge, even when the corresponding series of absolute values does not.
Its appeal lies in both simplicity and precision. A few checks on the size and limit of the terms can settle convergence, and the same assumptions provide a practical error bound for partial-sum approximations. Understanding the Leibniz test therefore pays off in pure analysis, numerical methods, and any setting where alternating signs appear naturally.
What Is the Leibniz Test?
The Leibniz test is a convergence criterion for an alternating series of the form sum (-1)^(n-1) b_n or sum (-1)^n b_n, where each b_n is positive. The test says that if the b_n eventually decrease monotonically and tend to zero, then the alternating series converges. This result is also known as the Leibniz criterion or the alternating series test.
It is important to note what the test does not claim. It does not guarantee absolute convergence, and it does not say that a series diverges when its conditions fail. A series may still converge for other reasons. The Leibniz test is a sufficient condition for convergence, not a necessary one.
Formal Statement of the Alternating Series Test
Let {b_n} be a sequence of positive real numbers. If b_(n+1) <= b_n for all sufficiently large n and lim b_n = 0 as n approaches infinity, then the alternating series sum (-1)^(n-1) b_n converges. The same conclusion holds for sum (-1)^n b_n. The monotonic decrease need only hold eventually; finitely many initial terms do not affect convergence.
The intuition behind the proof is geometric. The even partial sums form an increasing sequence, the odd partial sums form a decreasing sequence, and the gap between corresponding partial sums is b_(n+1), which shrinks to zero. These two bounded, monotone sequences squeeze toward the same limit, so the full sequence of partial sums converges.
Why the Conditions Matter
The decreasing condition cannot be discarded. Consider a series whose positive magnitudes are 1/n for odd n and 1/n^2 for even n. The magnitudes tend to zero, but they do not decrease monotonically because the odd terms are much larger than the preceding even terms. The resulting alternating series diverges, showing that monotonicity is more than a technical convenience.
The limit condition is equally essential. If b_n does not tend to zero, then the individual terms of the alternating series do not tend to zero, and the series diverges by the nth-term test. When both conditions hold, however, the alternating structure prevents the partial sums from running away, producing convergence even in delicate cases such as the alternating harmonic series.
Applying the Leibniz Test in Practice
To apply the Leibniz test, first identify the alternating factor, usually (-1)^n or (-1)^(n-1). Then isolate the positive magnitude b_n and check two properties: that b_n is eventually decreasing and that its limit is zero. If both checks pass, the series converges. If either fails, the test is inconclusive, though a failure of the limit test often proves divergence directly.
The test is especially useful for series that are not absolutely convergent. For example, sum (-1)^(n-1)/n converges by Leibniz, but sum 1/n diverges. This distinction between conditional and absolute convergence is central to analysis, and the Leibniz test is often the quickest way to establish the conditional case.
Worked Examples and Error Bounds
For the alternating harmonic series, sum (-1)^(n-1)/n, set b_n = 1/n. The sequence 1, 1/2, 1/3, ... decreases to zero, so the Leibniz test proves convergence. Yet the harmonic series sum 1/n diverges, so the alternating harmonic series converges only conditionally. Its sum is ln 2, a classic result that surprises many students.
The Leibniz test also supplies an error estimate. If S is the exact sum and S_N is the Nth partial sum, then |S - S_N| <= b_(N+1). In the alternating harmonic series, for instance, the error after 100 terms is at most 1/101. This bound makes the test valuable for numerical approximation, because it tells you how many terms are needed to achieve a desired accuracy without knowing the exact sum.
Common Pitfalls and Comparisons
A common mistake is assuming that every alternating series converges. The series sum (-1)^n (1 + 1/n) alternates, but its terms approach 1 rather than 0, so it diverges. Another pitfall is checking only the limit and ignoring monotonicity. As shown earlier, terms can tend to zero without decreasing and still produce divergence in an alternating setting.
It is also important to separate the Leibniz test from tests for absolute convergence. The ratio test, root test, and comparison test may show that sum |a_n| converges, which implies convergence of sum a_n and permits rearrangement. If only the Leibniz test applies, the series may be conditionally convergent, and rearranging its terms can change the sum or even make the series diverge.
The Leibniz test remains a compact but powerful idea: alternating signs plus shrinking magnitudes create convergence, and the size of the next term controls the approximation error. Whether you are checking a homework series, designing a numerical algorithm, or exploring the strange behavior of conditionally convergent series, this test gives you a reliable first move. Keep it beside the other convergence tests, and let the structure of the series guide which tool you reach for next.