$$\sum_{n=2}^\infty \frac1{n\ln n}$$ diverges by the integral test and p-series.


$$\sum_{n=2}^\infty \frac1{n^{1.1}}$$ converges by p-series, and is greater than the first series, right? So, by the direct comparison test for series, the first series should converge.

I know that a divergent series can't be smaller than a convergent series, so this must be wrong somehow. I think that

$$\frac1{n\ln n} < \frac1{n^{1.1}}$$

must be wrong, but I don't know how to show it. From what I can tell, the first series is larger at first, but at a very large n, is much smaller. Thanks for the help!!

  • 3
    $\begingroup$ You're wrong: $n\log n=o\bigl(n^{1.1}\bigr)$, so $\dfrac1{n^{1.1}}=o\Bigl(\dfrac1{n\log n}\Bigr)$. $\endgroup$ – Bernard Oct 26 '17 at 23:10

Sorry to burst your bubble, but you're wrong about $$\frac{1}{n\ln(n)}\lt \frac{1}{n^{1.1}}$$ Notice that for positive integer $n$, this statement is equivalent to $$\ln(n)\gt n^{0.1}$$ First of all, this is false for $n=1,2,3$, but that doesn't matter, since we're looking at asymptotic domination here. So we should instead consider whether or not $$\ln(n)\gt n^{0.1}$$ is true for large $n$. If this is true for large $n$, then the value of the limit $$\lim_{n\to\infty} \frac{\ln(n)}{n^{0.1}}$$ should be $\infty$. However, by L'Hopital, $$\lim_{n\to\infty} \frac{\ln(n)}{n^{0.1}}=\lim_{n\to\infty} \frac{\frac{1}{n}}{0.1n^{-0.9}}$$ $$\lim_{n\to\infty} \frac{\ln(n)}{n^{0.1}}=\lim_{n\to\infty} \frac{10}{n^{0.1}}$$ $$\color{red}{\lim_{n\to\infty} \frac{\ln(n)}{n^{0.1}}=0}$$ Which disproves your claim that $$\frac{1}{n\ln(n)}\lt \frac{1}{n^{1.1}}$$

  • $\begingroup$ Thank you!! I was sure that the inequality was wrong but I didn't know why. Thank you for clarifying $\endgroup$ – J. Door Oct 26 '17 at 23:26
  • $\begingroup$ @J.Door No problem! $\endgroup$ – Franklin Pezzuti Dyer Oct 26 '17 at 23:28

Note the inasmuch as $\log(n)\le n-1<n$, and $\log(n^{0.1})=0.1\log(n)$, we see that

$$\sum_{n=2}^N \frac1{n\log(n)}\ge 0.1\sum_{n=2}^N \frac{1}{n^{1.1}}\tag 1$$

One cannot conclude from $(1)$ that the partial sums on the left-hand side converges (they do not) as a consequence of convergence of the right-hand side of $(1)$.


Perhaps it is more clear to consider $n=e^{x_n}$ so that we get




Which one converges to zero faster? Hopefully you realize exponential growth dominates linear growth.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.