Did you know? A geometric series can add, subtract, add, subtract, and still settle towards one exact total. The alternating numbers 6, −3, 1.5, −0.75, … look as though they will keep pulling the total in opposite directions. Yet their partial sums approach four. The reason is a negative common ratio whose magnitude is smaller than one.
For JC1 H2 Mathematics learners in Bukit Timah, this is a valuable distinction in the sequences-and-series strand of the 2027 Singapore-Cambridge A-Level Mathematics 9758 syllabus. Students may learn S∞ = a/(1 − r) but forget the condition attached to it. They may also see a negative ratio, worry about the signs, and decide incorrectly that an infinite sum cannot exist.
This JC1 H2 Mathematics Tutor Bukit Timah guide develops one idea in depth: how the sign and size of a geometric ratio influence terms, partial sums and convergence. It includes two worked examples, a counterexample that does not converge, the full finite-sum method and a short self-test. For identifying individual terms from running totals, see the separate JC1 H2 guide to Sₙ and uₙ.
Quick answer: negative changes the signs; magnitude controls convergence
In a geometric progression, each term is found by multiplying the preceding term by one constant r. The nth term is uₙ = arⁿ⁻¹, where a is the first term.
If r < 0 and a is non-zero, the terms alternate in sign. If, in addition, |r| < 1, their magnitudes shrink towards zero and the infinite geometric series converges. Its sum is S∞ = a/(1 − r).
- r = −1/2: signs alternate, magnitudes halve, so a non-zero geometric series converges.
- r = −1: signs alternate, magnitudes stay constant, so the partial sums do not converge for a non-zero first term.
- r = −2: signs alternate, magnitudes double, so the series does not converge for a non-zero first term.
- r = 1/2: signs remain the same when a is positive, magnitudes halve, and the geometric series converges.
- r = 0: only the first term may be non-zero; the series has finite total a.
These statements assume a non-zero first term when discussing divergence. The trivial all-zero series has sum zero regardless of r.
Worked example 1: 6 − 3 + 1.5 − 0.75 + …
The first term is a = 6. The second term divided by the first is −3/6 = −1/2. Check the next ratio: 1.5/(−3) = −1/2. So this is a geometric progression with common ratio r = −1/2.
Write the first four partial sums:
- S₁ = 6.
- S₂ = 6 − 3 = 3.
- S₃ = 6 − 3 + 1.5 = 4.5.
- S₄ = 6 − 3 + 1.5 − 0.75 = 3.75.
The sums jump from above four to below four, then above, then below, each time getting closer. This is the essential picture of an alternating, convergent geometric series.
Because |r| = 1/2 < 1, the infinite sum exists. Calculate S∞ = a/(1 − r) = 6/[1 − (−1/2)] = 6/(3/2) = 4.
Do not turn 1 − (−1/2) into 1 − 1/2. Subtracting a negative is adding. The denominator is 3/2, not 1/2. That tiny sign error would incorrectly produce a sum of 12.
Why the sum can exist even though terms switch signs
Consider what happens in pairs. The first pair is 6 − 3 = 3. The next pair is 1.5 − 0.75 = 0.75. The following pair is 0.375 − 0.1875 = 0.1875. Each pair makes a positive addition, but the contributions rapidly become smaller.
Grouping into valid consecutive pairs illustrates why the partial sums can approach a finite total. However, always use the standard convergence condition before changing the grouping of an infinite series. We are safe here because a geometric series with |r| < 1 converges absolutely.
It is the shrinking magnitude, not a positive sign, that allows the sum to stabilise. A sequence of alternating terms that never shrink can keep a running total oscillating forever.
The finite-sum formula makes the limit visible
For a geometric progression with ratio r ≠ 1, the sum of the first n terms is Sₙ = a(1 − rⁿ)/(1 − r).
For our example, Sₙ = 6[1 − (−1/2)ⁿ]/(3/2) = 4[1 − (−1/2)ⁿ]. When n is odd, (−1/2)ⁿ is negative, so Sₙ is just above four. When n is even, (−1/2)ⁿ is positive, so Sₙ is just below four.
The difference from the limit is Sₙ − 4 = −4(−1/2)ⁿ. Its magnitude is 4(1/2)ⁿ, which approaches zero. This is an exact mathematical explanation of the alternating approach we saw in the first four partial sums.
For example, at n = 3, S₃ = 4[1 − (−1/8)] = 4(9/8) = 4.5. At n = 4, S₄ = 4(1 − 1/16) = 3.75. The formula and arithmetic agree.
Worked example 2: 12 − 4 + 4/3 − 4/9 + …
The first term is a = 12 and the common ratio is r = −1/3. Since |r| = 1/3 < 1, the infinite sum exists.
Compute S∞ = 12/[1 − (−1/3)] = 12/(4/3) = 9. The first few partial sums are 12, 8 and 28/3 (approximately 9.3333). They alternate around nine.
After the fourth term, S₄ = 12 − 4 + 4/3 − 4/9. Put the fractional parts over nine: 4/3 − 4/9 = 12/9 − 4/9 = 8/9. Therefore S₄ = 8 + 8/9 = 80/9, approximately 8.8889. This is below nine, just as the alternating pattern suggests.
Exact fractions are easier to verify than hurried decimal rounding. Comparing consecutive partial sums with nine shows how the running total alternates around its limit while getting closer to it.
Why r = −1 does not work
Now consider a = 6 and r = −1. The terms are 6, −6, 6, −6, … . The partial sums are 6, 0, 6, 0, … . They do not settle towards a single number.
The fixed expression a/(1 − r) would give 6/2 = 3 if used blindly. But the infinite-sum formula is invalid here, because |r| = 1, not less than one. The resulting number three is not the limit of the partial sums.
The error is not merely using the wrong denominator. It is forgetting to ask whether the sum exists before calculating it. Convergence is a mathematical condition, not a small-print note under a formula.
Why r = −2 fails even more dramatically
Take a = 2 and r = −2. The first terms are 2, −4, 8, −16, 32, … . Their magnitudes double rather than shrink. The partial sums are 2, −2, 6, −10, 22, … and move farther apart.
No finite number can be their limit. Although a/(1 − r) would formally give 2/3, that expression is not an infinite sum for this divergent series. It is merely the result of substituting into a formula outside its permitted conditions.
The quickest check is |r|. If |r| ≥ 1 and a ≠ 0, the infinite geometric series does not converge. This covers ratios greater than one, less than minus one, and the borderline ratios one and minus one.
Geometric sequence versus geometric series
A sequence is the ordered list of terms: 6, −3, 1.5, −0.75, … . A series adds those terms: 6 − 3 + 1.5 − 0.75 + … . The sequence of terms may approach zero, while the sequence of partial sums approaches a different number—four in our main example.
This distinction is the reason uₙ and Sₙ have different meanings. A student who says “the sum to infinity must be zero because the terms go to zero” has mixed up term behaviour with total behaviour.
When a non-zero geometric series has |r| < 1, its individual terms approach zero, but the partial sums approach a/(1 − r). For a general series, terms approaching zero are necessary but not enough on their own to prove convergence.
Why the negative common ratio is not a common difference
In an arithmetic progression, the difference between adjacent terms is constant. In our sequence 6, −3, 1.5, −0.75, the differences are −9, 4.5 and −2.25—certainly not constant.
The ratios, however, are all −1/2. That is what makes the sequence geometric. Students sometimes select an arithmetic sum formula because they see the numbers becoming smaller. But “decreasing in magnitude” is not the same as “subtracting the same amount.”
Before using any formula, write the first two adjacent differences and ratios. Then choose the rule that is actually constant. That small diagnostic prevents entire solutions from going down the wrong branch.
Practice: five questions with answers
Question 1. Find the common ratio of 8, −4, 2, −1, … . Answer: r = −1/2. Each term is multiplied by −1/2.
Question 2. Find the sum to infinity of that series. Answer: |r| = 1/2 < 1, so S∞ = 8/(1 + 1/2) = 16/3.
Question 3. Does 5 − 5 + 5 − 5 + … have an ordinary sum to infinity? Answer: No. Its partial sums alternate between five and zero; r = −1 fails |r| < 1.
Question 4. Does the geometric series with a = 3 and r = −3 converge? Answer: No. Its terms grow in magnitude and |r| > 1.
Question 5. A student writes S∞ = 10/(1 − r) for a geometric series but has not found r. What essential question must come first? Answer: Determine the common ratio and verify |r| < 1, apart from the trivial all-zero case.
The first three questions distinguish ratio, convergence and sum. A learner who repeatedly forgets the condition should write “|r| < 1?” beside every proposed infinite sum until the check becomes automatic.
A tutoring repair plan built around one error
A JC1 tutor might start with three short sequences: one with r = 1/2, one with r = −1/2, and one with r = −2. Ask the student to calculate five terms and four partial sums for each. This creates three visibly different stories: steadily approaching, alternating while approaching, and alternating while moving away.
Then introduce the formula only for the first two sequences. The student must explain why the third example is excluded. The final step is to have the learner derive Sₙ = a(1 − rⁿ)/(1 − r) by subtracting rSₙ from Sₙ and checking the resulting expression.
The strongest learning outcome is not remembering that negative ratios alternate. It is being able to say, “The sign makes the terms alternate, but the size of |r| determines whether the total can converge.”
Frequently asked questions
Does a negative geometric ratio always mean divergence?
No. A negative ratio such as −1/2 produces an alternating but convergent infinite geometric series when the first term is finite. Its magnitude is less than one.
Does an alternating series always converge?
No. The terms of 6 − 6 + 6 − 6 + … alternate but do not decrease in magnitude, and the partial sums fail to converge.
Why is S∞ sometimes smaller than the first term?
Later negative terms can subtract from the first term. In 6 − 3 + 1.5 − 0.75 + …, the limiting total is four, smaller than the initial six.
Can I use the formula if r = 1?
The finite geometric sum becomes Sₙ = na when r = 1. For a non-zero a, the infinite series diverges. Do not apply a/(1 − r), whose denominator is zero.
Is this relevant to current H2 Mathematics?
Yes. The 2027 official H2 Mathematics syllabus 9758 includes finite and infinite geometric series and the condition for convergence. School sequences of teaching can differ.
Where to continue
Read how to recover the nth term from Sₙ and recurrence sequences with fixed points for two other distinctions involving terms and totals. Explore graph transformations with modulus bars and the Bukit Timah H2 Mathematics overview for more JC foundations.
The eduKateSG small-group Mathematics teaching reference shows a diagnostic-first approach in a Secondary 1 programme. JC1 enquiries need their own confirmation through the Bukit Timah JC1 tuition page.
You can remember a negative geometric ratio in one sentence: the sign makes the total zigzag; the magnitude tells you whether those zigzags calm down.