Mistake Master

Radius and Interval of Convergence of Power Series BC only

Applying the ratio test to $\sum c_n(x-a)^{n}$ with $x$ fixed gives $L = \left|x-a\right|\lim\left|\frac{c_{n+1}}{c_n}\right|$, and solving $L < 1$ produces the radius directly, with $R = \infty$ when the coefficient ratio tends to zero, as for $e^{x}$, $\sin x$ and $\cos x$, and $R = 0$ when the series converges only at its centre. Substituting $x = a \pm R$ makes $L$ exactly $1$, because $\left|x-a\right| = R$ and the coefficient ratio tends to $\frac{1}{R}$, so the endpoints are precisely where the ratio test returns the inconclusive case 10.8 showed carries no information. Each endpoint must therefore be substituted and tested as its own numerical series, using a $p$-series or comparison argument, the alternating series test, or the nth term test, and the two endpoints are independent of each other.

All four combinations of endpoint behaviour occur, so no bracket can be assumed: $\sum x^{n}$ converges on $(-1, 1)$ with both endpoints failing the nth term test, $\sum \frac{x^{n}}{n}$ on $[-1, 1)$ since $x = -1$ gives the alternating harmonic series and $x = 1$ the harmonic series, and $\sum \frac{x^{n}}{n^{2}}$ on $[-1, 1]$ since both endpoints give a convergent $p$-series. Those three share a radius of $1$ and differ only by a power of $n$, which shows the radius and the endpoint behaviour to be independent. The radius is a single number and the interval is a set with two decided endpoints, so answering with one when the other was asked for is answering a different question; a complete interval answer contains the ratio test, the radius, a separate test at each endpoint, and the brackets those tests determine.

THE INTERVAL OF CONVERGENCE, CENTRED AT a. CONVERGES DIVERGES DIVERGES a − R a a + R THE RATIO TEST GIVES R. AT BOTH CIRCLES IT RETURNS L = 1. SO EACH ENDPOINT IS ITS OWN CONVERGENCE PROBLEM. THEY ARE INDEPENDENT, AND THEY OFTEN DISAGREE.
The circles are drawn undecided on purpose: the ratio test that produced $R$ returns $L = 1$ at both of them, so nothing about the shading determines whether either belongs to the interval. Substituting the endpoint gives a numerical series, and that series has to be tested with the machinery of 10.3 through 10.9.
FOUR OUTCOMES. ALL FOUR OCCUR, SO NONE CAN BE ASSUMED. LEFT END RIGHT END THE INTERVAL A SERIES THAT DOES IT DIVERGES DIVERGES (−1, 1) SUM OF x^n CONVERGES DIVERGES [−1, 1) SUM OF x^n/n DIVERGES CONVERGES (−1, 1] THE MIRROR OF ROW 2 CONVERGES CONVERGES [−1, 1] SUM OF x^n/n² ALL THREE EXAMPLES HAVE RADIUS 1.
The three named series differ only by a power of $n$ in the denominator and share a radius of $1$, so the radius carries no information at all about the brackets. Row two is the standard case: at $x = -1$ it is the alternating harmonic series and converges, at $x = 1$ it is the harmonic series and does not.

The work

3 ways in · any order
Lesson
Radius and Interval of Convergence of Power Series

Finds the radius of convergence with the ratio test, then shows that the same test returns its inconclusive value at both endpoints by construction, so each endpoint is a separate convergence problem needing its own test. Sets out the four interval shapes and gives a standard series realising each, all with the same radius.

Skill check · 10 scenarios
Diagnostic
10-item topic check

Ten items on radius and interval of convergence: applying the ratio test to find the radius, substituting each endpoint and testing the resulting numerical series separately, choosing the correct brackets, and distinguishing a radius from an interval.

Not yet available · 10 items
Targeted Practice
Drill a single misconception

Pick one of the failure modes you missed and drill it on its own. The round is adaptive: two correct in a row clears the misconception and moves you to the next.

Take the diagnostic to identify your misconceptions