Did you know? Two lines on a chart can start at the same number, 100, even when one originally represented 40 pupils and the other represented 200. The chart has not made the groups equal in size. It has deliberately reset each group’s starting value so we can compare relative change. Confusing that starting marker with a headcount changes the answer to the wrong question.
An index chart with a base year of 100 converts each original value into a percentage of its own chosen base-period value: index = current value ÷ base value × 100. A reading of 125 means that series is 25% above its base, not that it contains 125 people, dollars or books. If different series share the same base date and comparable definitions, their indexed movements can help readers compare growth even when original sizes differ.
The activity-club numbers here are invented and independently checked teaching examples, not actual Singapore school records or claims about pupil outcomes. The explanation is useful for class projects, reading economic data, interpreting education dashboards and asking whether a published chart compares like with like.
The first learner diagnostic: what does 100 actually stand for?
Ask a learner to complete this sentence: “On this chart, 100 means _____.” If the answer is “100 pupils,” pause. The correct answer is “the base-period level for that particular series.” A second question sharpens the test: if two lines both begin at 100, must the clubs have the same original membership? No. Different starting quantities can each be normalised to 100.
The axes and footnotes matter. Confirm that the vertical axis says something such as “Index, 2024=100,” rather than “students,” “$” or “percentage points.” Read whether the base is a single year, a monthly average or a group of years. An index is not automatically an inflation measure: classroom attendance, book loans or exercise completion counts can also be indexed, provided the units and definitions are coherent.
Worked example: one small club and one large club
A fictional school has two independent activity programmes. Its invented annual participant counts are: Science 40 in 2024, 50 in 2025 and 60 in 2026; Art 200 in 2024, 210 in 2025 and 240 in 2026. Treat each number as a consistent annual count under an unchanged participation rule, with no duplicate counting inside each programme. These assumptions are part of the demonstration.
Set 2024 to 100 for each programme. Science: 40÷40×100=100, 50÷40×100=125, 60÷40×100=150. Art: 200÷200×100=100, 210÷200×100=105, 240÷200×100=120. The check is reversible: Science 2026 index 150 means 1.5×its starting 40=60. Art 2026 index 120 means 1.2×its starting 200=240.
On the index chart Science’s line climbs from 100 to 150; Art’s climbs from 100 to 120. Science grew 50% relative to its own base, whereas Art grew 20% relative to its own base. Yet Art still has 240 participants, four times Science’s 60 in 2026. The index compares rates of growth, not absolute group sizes.
The actual changes also differ in an important way: Science gained 20 participants (60−40) and Art gained 40 (240−200). The larger absolute gain belongs to Art, even though Science’s proportional rise is steeper. A careful explanation identifies which question is being asked: more additional people, faster proportional growth, or larger final total.
The crucial distinction: index points are not percentage change
Between 2025 and 2026 the Science index goes from 125 to 150. The difference is 25 index points. It is not a 25% rise during that interval, because 2025 was already above the base. To calculate the 2025-to-2026 percentage change, divide by the earlier index: (150−125)÷125×100=20%. The raw-value calculation agrees: (60−50)÷50×100=20%.
Art goes from 105 to 120. That is 15 index points, equivalent to (120−105)÷105×100≈14.29% growth from 2025 to 2026. The raw counts confirm it: (240−210)÷210×100≈14.29%. Saying “the chart rose by 15%” would be wrong without the correct starting denominator.
This is the same distinction explained in the US Bureau of Labor Statistics’ technical notes on calculating index changes: an index-point difference and a percent change between periods are different quantities. The arithmetic works in education contexts as well, even though a classroom chart is not an official economic index.
What changes when the base year changes?
Suppose a student redraws the Science chart using 2025, not 2024, as the 100 reference. The new index values are 40÷50×100=80 for 2024, 50÷50×100=100 for 2025 and 60÷50×100=120 for 2026. Art with 2025=100 becomes approximately 95.24, 100 and 114.29. The plotted labels change, but the real participants do not.
Crucially, Science’s 2025-to-2026 growth remains 20%, and Art’s remains approximately 14.29%. Re-basing changes the reference marker, not the underlying relative change between the same two actual observations. The BLS guidance on rebasing demonstrates this property with official price-index series; our school arithmetic is a simpler illustrative application.
Never place two independently indexed series with different undisclosed base dates on one graphic and claim they are directly comparable. A Science series based at 2024=100 and an Art series based at 2025=100 do not have the same starting question. Either rebase both appropriately or explain the incompatible references.
Read the chart as a sequence of checks
- What quantity was originally measured? Participants, books, minutes, prices or something else? An index alone hides the original unit.
- What base period was set to 100? A particular year, month or multi-year average?
- Does each series use the same base and a consistent counting rule? If not, qualify the comparison.
- What question do I need to answer? Base-to-current relative growth, one-year growth, absolute change or final size?
- Can I reverse the index? Multiply index÷100 by the known base value to recover the current value, allowing for rounding.
- Do any values show an extraordinary jump? Check whether the definition or population changed before celebrating the change.
Guided practice: explain a line at 80
A fictional reading scheme records 75 books borrowed in its base month and 60 in a later month. Calculate the index: 60÷75×100=80. Its line is 20 index points below the base of 100 and its count is 20% lower than the base: (60−75)÷75×100=−20%. Ask a learner to explain the minus sign in words, then multiply 0.8×75 to recover 60 books.
Now alter only the later count to 90. The index becomes 90÷75×100=120, or 20% above the same base. If a learner responds “120 books” instead of “an index of 120 representing 90 books,” the diagnostic has exposed a confusion between an index number and a physical count.
Independent practice: compare two series without ranking children
Two invented school resource measures share a common base period. For Measure A, the base count is 80 and the current count is 96. For Measure B, the base count is 250 and the current count is 275. Find both indexes, name which increased faster relative to its own start, name which gained more units, and explain why the answers can be different. Work without looking at the check below.
Answer and audit: A’s index is 96÷80×100=120, a rise of 20%; B’s index is 275÷250×100=110, a rise of 10%. A grew faster proportionally. A gained 16 units, but B gained 25, so B had the larger absolute gain. Reverse each answer: 80×1.20=96 and 250×1.10=275. These results describe the fictional series, not relative quality or student achievement.
Misconceptions worth correcting early
- “100 means 100 things.” It means the reference value has been normalised to 100; it may represent many or few things.
- “150 means 150% growth.” From a base of 100, it is 50% growth, or 150% of the original level.
- “A 20-point rise means a 20% rise.” That is true from index 100 to 120, but false from 150 to 170, which is a 13⅓% rise.
- “A higher index means a larger group.” It can mean a faster relative rise in a much smaller group.
- “Index values tell us what caused the change.” They report numerical movement, not whether teaching, recruitment, a data-definition change or chance produced it.
- “A base of zero is acceptable.” The simple formula divides by the base value; a zero base makes this ratio undefined and requires a different reporting approach.
When a useful index becomes misleading
A student might construct an index of reading time, but one group’s base period was a short holiday week and another’s was an unusually busy examination week. Their relative trends would be mathematically correct yet difficult to compare fairly. Similarly, if enrolment rules or the set of participants change between years, a larger index need not prove that individual pupils improved.
A teacher should state the unit, sample definition, missing-value treatment, reference period and whether measurements are comparable. Do not fill absent months with zero simply to make the line neat. When the base count is very small, a change of one or two cases may create a dramatic percentage rise; report the original counts alongside the index to keep its size in perspective.
For families and teachers: make the learning observable
A parent can use two fictional allowance or reading-time sequences instead of comparing real children. Ask the learner what was set to 100, have them rebuild the indexed numbers, and then ask them to state what the chart cannot tell them. A teacher can test transfer with an unfamiliar chart of school equipment use, using the same three checkpoints: correct base, correct percentage-change denominator and a correctly limited claim.
Keep a short record across sessions: does the learner identify the base without prompting, convert an index back to its raw value when given the base count, and explain a change between non-base periods? Those behaviours show understanding more reliably than merely reproducing the formula from memory.
Frequently asked questions
Can an index fall below 100?
Yes. An index of 85 means the current level is 85% of the base level, or 15% below it, assuming a positive base and an ordinary ratio index.
Does 200 mean twice the base value?
Yes for the simple positive-base index used here. It is 200% of the base, equivalent to a 100% increase from that base—not a 200% increase.
Why do official statistical index calculations look more complicated?
Published indices may combine many items, weights, adjustments and methodological rules. Our single-series example teaches how to interpret a normalised reference and subsequent changes; it is not a complete recipe for constructing an official composite price index.
May I compare index lines that were set to 100 in different years?
Not as though they have the same reference. Rebase to a common period where methodologically sound, or keep the different bases explicit and limit your claim.
The core aim: distinguish the picture from the quantity
For related practice, see the Core Aim guide to why group averages cannot prove every pupil improved and combining class averages with different group sizes. The larger learning route appears in the eduKate Learning System and the Clementi Mathematics reference. Learning to name the denominator and check a result is useful far beyond one graph.