Measure geometric decay in a complicated term

Try

Consider aₙ=3ⁿ/n!. Computing large factorials directly is cumbersome, but dividing consecutive terms cancels nearly everything: aₙ₊₁/aₙ=3/(n+1). Predict what this says about how the far tail compares with a geometric series.

Predict

For aₙ=(2n/(3n+1))ⁿ, taking an nth root removes the outer power. Predict the limiting effective ratio. Then consider 1/n and 1/n²: their consecutive-term ratios both approach 1. A test based only on that limit must leave at least some cases undecided.

Experiment

Compute the simplified ratio or root before evaluating a limit.

Term	Ratio or root	Limit	Conclusion
3ⁿ/n!	3/(n+1)	0	absolute convergence
n!/5ⁿ	(n+1)/5	infinity	divergence
(2n/(3n+1))ⁿ	nth root 2n/(3n+1)	2/3	absolute convergence
1/n²	ratio (n/(n+1))²	1	test inconclusive

For the first sequence, find an index after which its ratio is at most 1/2. From that index onward, each term is bounded by a fixed multiple of successive powers of 1/2. Write this bound explicitly rather than describing the terms as merely “small.”

Understand

The ratio test uses the limit L of |aₙ₊₁/aₙ| when the ratios are eventually defined. If L<1, the series converges absolutely. If L>1, including an infinite limit, terms eventually grow in magnitude rather than tending to zero, so the series diverges. At L=1 the test is inconclusive. Zeros or nonexisting ratio limits may require a different test or a more general theorem; do not divide by a zero term.

The root test uses the limit L of |aₙ|^(1/n). Again L<1 gives absolute convergence and L>1 gives divergence; L=1 is inconclusive. A general limsup version covers cases without an ordinary limit, but this unit uses ordinary limits unless explicitly stated. Absolute values are essential when original terms change sign.

For the convergence argument, choose q strictly between L and 1. Eventually the root inequality gives |aₙ|≤qⁿ directly. For the ratio version, repeatedly applying an eventual ratio bound gives |a_N+k|≤|a_N|q^k. Both reduce the problem to geometric comparison. This explains why early large terms do not prevent convergence and why the method addresses tails rather than any fixed finite sum.

For n!/5ⁿ, the ratio eventually exceeds 2, so terms grow without bound. In contrast 3ⁿ/n! eventually has a ratio below any prescribed positive number, demonstrating that factorial growth in the denominator can overcome a fixed exponential numerator. The location of the factorial matters; naming a factorial alone does not determine convergence.

The boundary limit 1 genuinely contains both behaviors. The ratio for the harmonic series is n/(n+1); for the convergent p-series 1/n² it is its square. Both tend to 1. Their nth roots also tend to 1. An inconclusive result is not evidence of divergence, nor a reason to repeat the same test indefinitely. Use comparison, integral or alternating arguments instead.

Apply

In practice, simplify algebraically before taking the limit. State the absolute-value test and distinguish convergence from absolute convergence. Supply two contrasting limit-one examples and justify them with another theorem. In the proof task, show the geometric bound that turns a numerical limit into a convergence guarantee.

Continue

Next let the terms depend on a variable x. Ratio and root tests will determine an open convergence interval, while its endpoints need their own numerical-series tests.

Try the mixed root-test family with oscillating ratios, then choose related tasks from the university calculus practice guide. This public practice records reasoning and solution exposure; it does not award independent proficiency.