Mathematics Guides That Build Reasoning on Firm Foundations

Mathematics is built like a staircase. Quadratic equations in senior high rest on factorising from junior high, factorising rests on multiplication and negative numbers, and all of it rests on the fractions learned in primary school. A student stuck on one chapter is often carrying a gap from two or three levels below. Repeating the current chapter over and over rarely helps, because the footing underneath is what needs attention.

The way to find that footing is a backwards diagnostic. Prepare about 10 mixed questions from material two or three levels below the current class: fraction operations, negative integers, two-digit multiplication, ratio and simple linear equations. Work through them without a calculator and without a tight time limit. Each wrong answer is traced further back to more basic material until the last topic that is truly secure turns up. That is the starting point.

Once the starting point is clear, one topic is studied until the learner can get 8 out of 10 questions right, and only then does the next topic begin. The principle comes from Benjamin Bloom's research on mastery learning: prerequisites settled first make what follows far lighter. Where the gaps run deep, 8 to 12 weeks of steady practice is a reasonable picture, and every learner moves at a different pace.

You can tell the diagnosis was right when questions at the current level start to make sense, because each step can be traced back to an idea already known. If a question still feels foreign, there is usually one more gap underneath waiting to be found.

In Indonesia's Kurikulum Merdeka, multiplication is introduced in Phase A, around grade 2, as repeated addition, and deepened in Phase B in grades 3 and 4. From then on it is a working tool for division, fractions, LCM and GCF, and word problems. A grade 5 student still counting 7 x 8 on their fingers spends working memory on small sums, and the longer steps of a problem easily get lost.

Learning the table becomes much lighter when patterns come first. Times 2 means doubling, times 4 means doubling twice, times 10 means adding a zero, and times 5 has units that alternate between 5 and 0, like the numbers on a clock face. For 9 there is the finger method: hold up ten fingers, fold down finger n, and the fingers to the left give the tens while those to the right give the units. The commutative property covers the rest, because 7 x 8 equals 8 x 7, so the 100 cells of the table shrink to 55 unique pairs.

The few pairs that still feel heavy, such as 6 x 7, 7 x 8 and 8 x 6, are split into parts already known. For example, 8 x 7 becomes 8 x 5, which is 40, plus 8 x 2, which is 16, giving 56. Practice stays short, about ten minutes a day with cards or dice, and without a strict stopwatch, since timed tests leave some students tense enough to forget patterns they already understand.

The skill is in place when products come out without recounting from zero, and when a student who forgets one fact can rebuild it from a neighbouring pair they remember.

Formulas memorised raw evaporate quickly and get mixed up, especially when a question changes shape slightly. A formula you have derived yourself lasts much longer, because it is stored as a chain of reasons you can follow again. Dunlosky and colleagues' review of study techniques ranks this kind of active understanding well above rereading and highlighting notes.

Take three examples. The area of a triangle is half the area of a rectangle with the same base and height, which gives one half times base times height. A trapezium is cut into two triangles of equal height, and their areas are added. The quadratic formula comes from completing the square on ax² + bx + c = 0. And from Pythagoras' theorem on a right triangle with hypotenuse 1, the identity sin²x + cos²x = 1 follows directly.

In practice: watch a formula being derived once, derive it again by hand without looking, then write the meaning of each symbol in your own words. Related formulas go onto a map, for plane shapes or trigonometry, say, with parent formulas at the top and the ones derived from them underneath. The map cuts down how many formulas genuinely need memorising, since most can be built from a handful of parents.

A formula is understood when it can be derived again without notes and explained smoothly to someone else. Wherever the explanation stumbles shows exactly what needs another look.

Dunlosky and colleagues reviewed ten popular study techniques and found two with the strongest evidence across subjects and ages: spaced practice and retrieval practice. For mathematics, that means 20 to 30 minutes every day stays in memory better than three hours piled into one weekly session. Questions are attempted from memory first, and the worked solution is checked afterwards. Rereading notes and highlighting fall among the low-utility techniques.

The third technique is mixing question types once the basics of a topic are understood. Rohrer and colleagues showed that interleaved practice trains the ability to choose a method, something that never gets exercised when 20 questions of one type are done in a row. Mixed sessions feel harder, and that feeling is part of how the learning happens. One session might weave together linear equations, ratio and the area of plane shapes.

The companion to all this is an error notebook. Every wrong answer is recorded with one sentence about where it went wrong, then labelled by its source: the idea itself was shaky, the arithmetic went astray, or the question was read wrongly. The sorting matters because each kind has its own remedy. A concept error sends you back to the material, a calculation slip means slowing down and writing every step, and a misreading means strengthening the understanding stage. A few days later, old mistakes are reworked without looking at the solution.

The habit is working when questions once answered wrongly come out right after a gap, and the most frequent kind of error in the notebook starts to appear less often.

HOTS (Higher Order Thinking Skills) questions ask for pattern analysis, a chosen strategy and the reasoning behind an answer. One question can have several valid routes, and what earns credit is the soundness of the reasoning. Questions of this kind appear in primary-level olympiads, in the Mathematical Reasoning and Quantitative Knowledge subtests of Indonesia's UTBK university entrance test, and in the numeracy items of Indonesia's national assessment (Asesmen Nasional).

At primary level, the best-known official route is the Olimpiade Sains Nasional (OSN, called the Kompetisi Sains Nasional or KSN from 2020 to 2024), run by Indonesia's Ministry of Primary and Secondary Education, with Mathematics, Science and Social Studies as the primary subjects. Selection is tiered: school, then regency or city, province, and finally national. National winners may be selected again for IMSO, an international event for participants aged 13 or under. The material still rests on numbers, fractions, LCM and GCF, plane and solid shapes, measurement, and patterns and logic. Preparation usually begins 4 to 6 months before selection, using the official OSN primary question bank.

At any level the sequence for HOTS practice is the same: reasoning first, speed later. The early months go into strengthening core concepts and working one hard question all the way through, including hunting for a second route. The middle stage widens the variety of questions. As selection or the UTBK approaches, timed practice is introduced gradually, from generous limits down to something close to the real thing, while practising the habit of tackling the most familiar questions first.

For primary pupils, an olympiad works best when it still feels like a place to play with numbers and puzzles. At any level, readiness for HOTS shows in the habit of asking why a method works, and in being able to write that reason down in order.