Alternating Series and P-series convergence

Join the discussion
Ask a follow-up here, or get your own question answered by working scientists, mathematicians and engineers — people, not an autocomplete.
Real named experts · corrections over time · the nuance an AI answer skips
6 replies · 20K views
kylera
Messages
40
Reaction score
0
Alternating Series and P-series "convergence"

I couldn't resist trying out a pun. Anyway, onto the question:

Homework Statement


Test the series for convergence/divergence:
[tex]\sum^{\infty}_{n=1}\frac{-1^{n-1}}{\sqrt{n}}[/tex]


Homework Equations


Alternating Series Test and possibly p-series test...


The Attempt at a Solution


The expression for an alternating series goes as a(n) = (-1)^(n-1) * b(n). Having said that, it's obvious that b(n) is

[tex]\frac{1}{\sqrt{n}}[/tex]

which can be re-written as [tex]\frac{1}{n^\frac{1}{2}}[/tex]. But by the rules of the p-series, since 0.5 is obviously lesser than 1, I came to the conclusion that the series diverges. Instead, the answer states that the series does indeed converge. Can anyone help shed some light in this?
 
Physics news on Phys.org


Dick said:
Sorry, I don't get the 'pun'. Am I being thick?

I don't think you are since I don't get it either.
 


Well usually for the p-series test, the only examples with fractions I've ever seen is where the function is in the form

[tex]\frac{1}{f(n)}[/tex]
 


rock.freak667 said:
Well usually for the p-series test, the only examples with fractions I've ever seen is where the function is in the form

[tex]\frac{1}{f(n)}[/tex]

Sure, in fact, it applies only to series of the form 1/n^p or n^(p). I wasn't saying you were wrong. It's useful here if you want to discuss absolute convergence. It is useless for the given function.
 


The "p-series" only applies to positive series. That's why Dick says it is useful to discuss absolute convergence. If your series has both positive and negative terms then it may converge "conditionally". It will converge absolutely only if the series of absolute values converges.

By the "alternating series test", the series
[tex]\sum \frac{(-1)^n}{\sqrt{n}}[/tex]
converges.

By the "p- series test"
[tex]\sum \left|\frac{(-1)^n}{\sqrt{n}}\right|= \sum \frac{1}{\sqrt{n}}= \sum \frac{1}{n^{1/2}}[/tex]
does NOT converge and so the original series converges conditionally, not absolutely.