A logarithm may tell you that a growing quantity crosses a target after 6.116 days, but an Additional Mathematics question might ask for the first whole day on which the target is exceeded. These are not the same answer. First solve the real-valued exponential inequality, then apply the question’s strict or inclusive wording and the allowed time values.
For example, if A(n) = 200(1.12)^n is checked at the end of each whole day, the first day when A(n) exceeds 400 is day 7: day 6 gives about 394.76, and day 7 gives about 442.14. This guide shows how to make that decision reliably for growth, decay, doubling time, thresholds and word problems without rounding at the wrong stage.
Identify whether time is continuous or counted in whole steps
An exponential model often uses A(t) = A₀b^t, where A₀ is the initial amount and b is the multiplier per time unit. The symbol t can represent a continuous real-valued time or a discrete count of complete periods. The question must tell you which interpretation is intended.
If a sensor measures continuously, the model can cross a threshold at t = 6.116 days. That is a meaningful crossing instant. If the model is sampled only at the end of each full day, the first reported day beyond that threshold is n = 7.
It is inaccurate to call 6.116 “day 6” simply because its nearest integer is 6. At the end of day 6 the target is not yet exceeded. It is also inaccurate to report day 7 as the exact continuous crossing time. The result depends on the question.
Before taking logarithms, write the meaning of the independent variable: time since the initial reading, number of completed compounding periods, or a calendar day label. That small sentence determines how the final result should be communicated.
Build the multiplier correctly from a percentage
A quantity that grows by 12% every complete day is multiplied by 1.12 for each day. Thus a starting value of 200 becomes 200(1.12)^n after n completed days. The multiplier is 1 + 0.12, not 0.12.
If a quantity decays by 20% per day, 80% of its previous amount remains after each day. The multiplier is 0.80, not −0.20. A model with initial amount 1000 is therefore A(n) = 1000(0.8)^n.
Students should distinguish the growth or decay rate from the retained multiplier. A percentage is converted to a decimal before the multiplier is formed. Saying “12% growth” does not mean “multiply by 12” or “add 12 every day”.
A quick one-step calculation catches most errors. With 12% growth, day 1 must give 200 × 1.12 = 224. With 20% decay, day 1 must give 1000 × 0.8 = 800. If a proposed formula fails this elementary check, do not continue to logarithms until the model has been repaired.
Work a complete growth-threshold question
Suppose A(n) = 200(1.12)^n, with n a non-negative integer counting completed days. Find the first whole day when A(n) is strictly greater than 400.
Divide the desired inequality by the positive initial value 200. The result is (1.12)^n greater than 2. Because 1.12 is greater than 1, its exponential function increases, so taking logarithms yields n greater than ln 2 / ln 1.12.
The boundary is approximately 6.116255. The smallest integer greater than that boundary is n = 7. This is an integer decision following a mathematical inequality, not conventional rounding to the nearest whole number.
Check adjacent times. A(6) = 200(1.12)^6 ≈ 394.76, below the target. A(7) = 200(1.12)^7 ≈ 442.14, above the target. The check validates both the answer and the interpretation: the threshold is passed between the end-of-day 6 and end-of-day 7 readings.
Why rounding 6.116 to six is wrong
Ordinary rounding to the nearest integer answers the question “which integer is numerically closest to 6.116?” It does not answer “what is the smallest allowed integer satisfying n greater than 6.116?” Those are different operations.
For a strict lower bound n greater than a positive real number c, the first permitted integer is floor(c) + 1. Here floor(6.116…) + 1 = 7. For an inclusive lower bound n at least c, the first integer is the ceiling of c, provided n must be an integer.
If c is not itself an integer, these two rules happen to agree. They differ at an exact integer boundary, so students must retain the distinction rather than memorising “always round up” without reference to the inequality.
The final verification should be stated in words. At day 6 the amount has not exceeded the target. At day 7 it has. That sentence proves minimality: it shows that the chosen day succeeds and the immediately preceding whole day fails.
The difference between ‘exceeds’ and ‘at least’
The wording “exceeds 800” means strictly greater than 800, while “is at least 800” means greater than or equal to 800. The distinction matters when the model reaches the threshold exactly at a whole-number time.
Take A(n) = 100 × 2^n. At n = 3, A(3) = 800. Therefore the first day when the amount is at least 800 is day 3. But the first day when the amount exceeds 800 is day 4, when the amount becomes 1600.
If a learner blindly takes the ceiling of a logarithm result of exactly 3 for both questions, they will answer 3 twice and get the strict case wrong. The numerical model is identical; only the logical relation changed.
In school word problems, underline phrases such as “more than”, “not less than”, “at most”, “below” and “does not exceed”. Translate the wording into a relationship before performing any logarithmic manipulation. That translation is part of the mathematics.
The general growth-threshold method
For a model A(n) = A₀b^n with A₀ positive and b greater than 1, suppose the target T is greater than A₀. To find the first non-negative integer n with A(n) at least T, first solve b^n at least T/A₀.
Taking logarithms gives n at least ln(T/A₀)/ln b, because ln b is positive. The first integer candidate is the ceiling of that real number. If the requirement is strictly greater than T, the first integer candidate is the floor of that real number plus one.
There is a simple caveat: when the threshold is already satisfied at n = 0, the answer can be day 0 rather than a later day. The domain of n and the meaning of the initial observation are part of the question.
This formula is best understood as a boundary-and-check procedure, not an invitation to type a complicated expression and trust the displayed decimal. Determine the direction of the inequality, choose the first permissible integer and evaluate the adjacent steps.
The logarithm base does not change the boundary
To solve (1.12)^n greater than 2, we may write n greater than ln 2 / ln 1.12. We could also use common logarithms: n greater than log 2 / log 1.12. Both ratios are equal because of the change-of-base formula.
Students sometimes mix a natural logarithm in the numerator with a common logarithm in the denominator. That is not a legitimate change-of-base calculation and will generally give an incorrect boundary.
Use the same logarithm base in the numerator and denominator. The choice of base is a calculation convenience; the underlying time is determined by the exponential relationship itself.
Check reasonableness without a calculator. Since (1.12)^6 is a little less than 2 and (1.12)^7 is greater than 2, the boundary must lie between 6 and 7. Any logarithmic result of 0.6 or 61 should prompt investigation before selecting the day.
Exponential decay reverses an inequality when taking logs
Now consider A(n) = 1000(0.8)^n and ask for the first completed day when the amount is less than 500. Divide by 1000 to obtain (0.8)^n less than 0.5.
Take natural logarithms: n ln 0.8 is less than ln 0.5. Because ln 0.8 is negative, dividing by it reverses the inequality. Therefore n is greater than ln 0.5 / ln 0.8, approximately 3.106284.
The first whole day satisfying the condition is n = 4. Indeed A(3) = 512, which has not fallen below 500, while A(4) = 409.6, which has.
The sign reversal is not an optional detail. Decay has a multiplier between zero and one, so its logarithm is negative. Students who write the same inequality direction as for growth may conclude that earlier times satisfy a later-decay target, a contradiction exposed immediately by evaluating day 0 and day 1.
Decay does not make the amount negative
For positive A₀ and a decay multiplier b strictly between zero and one, A₀b^n remains positive at every finite real n. It approaches zero as n grows, but does not become negative through this simple model.
This is worth checking when a student interprets a negative logarithm or divides by ln b. A negative value of ln 0.8 does not mean the measured amount is negative. It reflects the fact that the exponential multiplier decreases with time.
The model may cease to be realistic over extremely long periods, and some real systems have background amounts or floors. Those features require a richer model, not arbitrary negative outputs introduced into the simple exponential expression.
In examination work, a fast reasonableness check is to compare the first few values with the story. Under 20% decay per day, values should be 1000, 800, 640, 512 and 409.6. An increasing sequence would indicate that the multiplier or interpretation is wrong.
Half-life is a boundary in a decay model
A half-life is the time required for a modelled quantity to fall to half its initial value under exponential decay. For A(t) = A₀b^t with 0 less than b less than 1, solve b^t = 1/2.
The real-valued half-life is t = ln(1/2)/ln b. For b = 0.8, it is about 3.1063 time units. If the question asks when a continuously changing amount is exactly half, that decimal time is meaningful.
If instead measurements occur at the end of complete days and the question asks for the first reported amount below half, the answer is day 4. If the sample is taken only at day boundaries, no observation occurs at the exact mathematical half-life.
Teach the distinction explicitly. “The half-life is approximately 3.106 days” and “the first whole-day reading below half occurs at day 4” can both be correct statements about the same model.
An initial threshold may already be satisfied
Suppose a model begins with 100 units and grows by 5% per day, but the question asks for the first day when the amount exceeds 80. The condition already holds at n = 0. Therefore the answer is the initial observation, day 0, if the question allows that label.
Blindly applying a logarithm formula without checking the starting value can produce an unhelpful negative boundary. A negative real crossing time might describe when the model would have crossed the target if extrapolated backwards, not the first permitted non-negative day.
The correct method begins by evaluating A(0) = A₀ against the required condition. Only if it fails should the learner solve for a later threshold crossing.
Also read the problem’s day numbering. Some descriptions call the starting amount “Day 1” rather than “Day 0”. In that case the mathematical index and the displayed calendar label may differ by one. Define the index explicitly before reporting the answer.
A threshold from a different starting day
Suppose a quantity grows by 10% per day and is known to be 500 units at the end of day 4. If n counts days after that reading, the model is A(n) = 500(1.1)^n. The exponent n = 0 refers to the end of calendar day 4, not the original beginning of the process.
If the amount must exceed 600, solve (1.1)^n greater than 1.2. The boundary is ln 1.2 / ln 1.1 ≈ 1.9129. The first integer n satisfying it is 2, meaning the end of calendar day 6 under this indexing.
Students can produce an apparently correct logarithm calculation and still lose a mark by reporting “day 2” when the question asks for the calendar day after a known day-4 reading. The model index and calendar label must be distinguished.
A short timeline with the known reading, n = 0, n = 1 and n = 2 prevents an off-by-one mistake. It also helps clarify whether amounts are measured at the start or end of each day.
Use a two-row final check, not a full recalculation
Once the boundary has been found, calculate the model value at the proposed first integer day and the immediately preceding day. The proposed day must satisfy the condition; the preceding day must fail it.
For A(n) = 200(1.12)^n and threshold 400, use n = 6 and n = 7. Their approximate outputs are 394.76 and 442.14. For A(n) = 1000(0.8)^n below 500, use n = 3 and n = 4, giving 512 and 409.6.
This two-row test proves the “first” part in a monotonic model. It is far stronger than reporting an integer derived from ordinary rounding and assuming the answer is self-evident.
Do not round the evaluated amounts so coarsely that both appear equal to the target. Keep enough intermediate precision to distinguish the side of the boundary, especially when the threshold is close to an exact day value.
Rounding near an exact boundary can mislead
Suppose an exact logarithm ratio is an integer, but a calculator displays 2.999999999 due to finite precision. If you treat that displayed number as genuinely smaller than 3 and apply a strict-inequality rule mechanically, you could select an invalid earlier day.
Use exact structure where possible. For A(n) = 100 × 2^n and target 800, observe 800/100 = 8 = 2³ exactly, so the boundary is n = 3. No decimal approximation is required to determine whether “exceeds” means day 4.
For other bases, compute enough precision and check the boundary values directly. A model evaluation at consecutive integer days is a more dependable final arbiter than the last decimal place of a logarithm shown on a screen.
This is a general lesson in Additional Mathematics: approximation is a tool, not permission to discard the exact logical relationship. Report the integer answer only after the intended inequality has been tested.
Exponential growth is not repeated simple addition
If an amount increases by 12% each day, the daily increase is calculated from the previous day’s amount. The sequence 200, 224, 250.88 and so on is multiplicative. Adding a constant 24 every day would instead model linear growth based on 12% of the original amount alone.
This distinction changes threshold times. A student using A(n) = 200 + 24n may obtain a plausible-looking day but has represented a different process. The first check is to ask whether the statement says “12% of the current amount” or “increases by 24 units per day”.
A fixed multiplier generates an exponential function; a fixed absolute difference generates a linear function. The question’s language should determine the model before logarithms are introduced.
If a mathematical model does not match the description, more accurate calculation will not repair the answer. Modelling is part of the reasoning being assessed, not just a preface to the equation-solving stage.
The first full period and the exact crossing time are complementary
Consider a continuous exponential model A(t) = 200(1.12)^t, where t can be any non-negative real time. It reaches 400 at t = ln 2 / ln 1.12 ≈ 6.1163 days, and exceeds 400 at times after that instant.
If the quantity is observed at integer t-values, the first recorded value above 400 occurs at t = 7. If a report asks when during the week the crossing happens, a statement such as “between the end of day 6 and the end of day 7” is more informative than an unexplained “7 days”.
This is not a contradiction between real and integer solutions. One identifies an instant on a continuous model; the other identifies the earliest allowed observation in a discrete sequence.
When giving a final sentence, repeat the time convention: “The first end-of-day reading that exceeds 400 is day 7.” That prevents a technically sound calculation from being interpreted as a claim about the exact crossing instant.
Distinguish reaching, exceeding and staying below
Word problems may use “reaches”, “exceeds”, “remains below” or “falls to or below”. These are logical statements with slightly different boundary conditions. For a monotonic model they can be solved systematically, but the inequality must match the wording.
With A(n) = 100 × 2^n and target 800, “first reaches at least 800” gives n = 3, “first exceeds 800” gives n = 4, and “last day still below 800” gives n = 2. Each refers to the same boundary but selects a different integer relation.
For decay A(n) = 1000(0.8)^n, “first falls below 500” gives n = 4. “Last day at or above 500” gives n = 3. The two answers are adjacent because the sequence crosses the threshold between those readings.
Write a tiny inequality before manipulating the model. In a long assessment, the five seconds spent translating the verb accurately can prevent a wrong final day despite correct logarithmic calculations.
A table can reveal a faulty model before algebra
Early values provide a simple independent check. For A(n) = 200(1.12)^n, the first three terms are A(0) = 200, A(1) = 224 and A(2) = 250.88. The increments are growing because a fixed percentage is applied to a larger amount each period.
For A(n) = 1000(0.8)^n, the first terms are 1000, 800 and 640. The amount decreases, but the absolute drop becomes smaller each period. This is consistent with a constant proportional reduction.
Such tables are not substitutes for logarithms when the threshold lies many periods away. They are model checks and help students see the trend that determines the inequality direction.
If the student solves an increasing sequence and concludes that later days fall below an already exceeded threshold, the table exposes the interpretation error. Algebra and numerical pattern recognition should support each other.
A second growth example with a new target
Suppose savings follow S(n) = 100(1.05)^n and the target is 170. The real boundary solves (1.05)^n = 1.7, so n = ln 1.7 / ln 1.05 ≈ 10.8757.
The first completed year or period at which the recorded value is at least 170 is n = 11, if n is restricted to non-negative integers. At n = 10, the model gives approximately 162.89; at n = 11, it gives approximately 171.03.
If instead the question asks for the precise continuous time when the model reaches 170, the appropriate answer is approximately 10.876 time units, subject to the requested accuracy.
The same algebra thus supports two final reporting conventions. Ask the student which answer is appropriate before looking at a marking scheme. Reading the words “first complete year” is as important as applying the logarithm law correctly.
How to handle a decreasing model with an upper target
Suppose a resource concentration follows C(n) = 600(0.9)^n, with daily end readings and a requirement to find the first day when the concentration is at most 300. The equation at the boundary is (0.9)^n = 0.5, giving n = ln 0.5 / ln 0.9 ≈ 6.5788.
Because the multiplier is less than one, later days have smaller values. The first integer with C(n) at most 300 is n = 7. At day 6, C(6) ≈ 318.86; at day 7, C(7) ≈ 286.98.
The inequality direction must be handled using the negative logarithm of the decay base. It is often safer for learners to solve the equality first, identify that the sequence decreases, then test neighbouring integer times.
This equality-boundary-plus-check method can be easier to understand than symbolically manipulating a string of inequalities while worrying about sign reversal. The final consecutive-day evaluation still determines the correct answer.
Avoid conflating a compound model with an exact forecast
The formula A(n) = A₀b^n assumes a constant multiplier for every period considered. Real populations, bank accounts, resource quantities and biological processes may change growth rates or face external constraints. A correct result is a conclusion under the stated model, not a guarantee that reality will follow it indefinitely.
Students should answer the mathematical question without inventing additional conditions. If a model assumes 12% daily growth, use that rate consistently unless the question explicitly introduces a change. But do not extrapolate a school exercise to assert a real financial outcome or health prediction.
When a question asks for a minimum time, it is reasonable to say “according to the model”. This phrase distinguishes mathematical calculation from empirical certainty.
Modelling discipline is important in Additional Mathematics: define quantities, use the relationship supplied, solve accurately, and interpret the result within the scope and limitations stated.
Common questions from students and parents
Should I always round the logarithm result up? Not automatically. Check whether the question asks for a whole period, whether the inequality is strict and whether the exact boundary is already an integer.
Why does a decay inequality reverse after taking logarithms? The logarithm of a base strictly between zero and one is negative; dividing an inequality by a negative number reverses its direction.
Can day 0 be the correct answer? Yes, if the initial reading already meets the target and the time indexing includes day 0.
Does “exceeds” mean “at least”? No. “Exceeds” is strict; “at least” includes equality.
Can I use common logs rather than natural logs? Yes, provided the numerator and denominator use the same base and all logarithm arguments are valid.
Is this question relevant to 2027 SEC Additional Mathematics? Exponential and logarithmic problem types should be checked against the student’s specific syllabus. The official G3 K341 and G2 K232 lists identify the corresponding documents. For the 2026 O-Level, check 4049.
Practice three threshold types independently
Task 1: A(n) = 100 × 2^n. Find the first integer n when A(n) exceeds 800. The answer is n = 4 because n = 3 gives exactly 800.
Task 2: A(n) = 1000(0.8)^n. Find the first integer n when A(n) is below 500. The answer is n = 4 because the values at n = 3 and 4 are 512 and 409.6.
Task 3: A(n) = 100(1.05)^n. Find the first integer n when A(n) reaches at least 170. The answer is n = 11, supported by the adjacent values 162.89 and 171.03.
For every task, write the starting amount, multiplier, inequality, real boundary, permitted integer and adjacent-value check. Then change one word—such as “exceeds” to “reaches at least”—and explain whether the answer changes. This is a much stronger test of understanding than memorising one rounding instruction.
Next steps for A-Math revision
For a fresh problem, begin with the question’s time convention and threshold wording. Build the exponential equation, use logarithms to find a real boundary and decide whether the answer should be a decimal instant or a whole-number observation. Verify consecutive integer values if the question asks for the first whole period.
For foundations, read Why Study Additional Mathematics | Use Logarithms to Find How Long Growth Takes and How to Improve Additional Mathematics | Logarithmic Equations and Domain Checks. To retain the technique, add a delayed problem to the weekly A-Math question queue.
Explore more in the Additional Mathematics archive or discuss suitable small-group tutorials with eduKate Singapore. The important achievement is not obtaining a decimal from a calculator; it is making the correct logical decision about the first allowed time. Properly taught kids shine a bright light into the future.