Truncating intermediate probability values, resulting in final answers that deviate from the required 4 d.p. standard.

Store intermediate calculation results in the calculator's memory registers (A, B, C, etc.) or write them down to at least 6 decimal places.

Neglecting to divide the variance or standard deviation by \( \sqrt{n} \) when standardising the sample mean under the Central Limit Theorem.

Always check whether the variable is an individual observation \( X \) or a sample mean \( \bar{X} \). For the latter, use \( \text{Var}(\bar{X}) = \sigma^2/n \).

Failing to write down the differential symbol dx or dt, or failing to change the limits of integration when performing integration by substitution.

Always write the differential step explicitly (e.g., \( du = g'(x)dx \)) and construct a small table to convert \( x \)-limits to \( u \)-limits before integrating.

Using the sample standard deviation \( s \) directly as the margin of error without dividing by \( \sqrt{n} \) when calculating Confidence Intervals.

Recall the confidence interval formula: \( \bar{x} \pm z_{\alpha/2} \left(\frac{s}{\sqrt{n}}\right) \). The term \( s/\sqrt{n} \) is mandatory.

Incorrectly defining the conditional probability space in Bayes' Theorem, especially missing complementary events in the denominator.

Draw a probability tree diagram immediately. Ensure that all branches leading to the target event are summed in the denominator.

HKDSE · Exam Tips

Mathematics M1 (Calculus and Statistics) Exam Tips

Comprehensive preparation package for HKDSE Mathematics Module 1 (Calculus & Statistics), featuring time-management breakdowns, examiner-verified marking traps in integration and probability, and a custom Trapezoidal Rule calculator program.

3 min readUpdated: Jun 21, 2026

Exam at a Glance

Papers

Total Marks

100

Time Limit

2h 30min

Question Types

Paper	Duration	Marks	Questions	Weighting	Question Types
Module 1 (Calculus and Statistics)	2h 30min	100	12	100%	Section A (Short Questions), Section B (Long Questions)

Grade Scale

5**5*54321

Calculator Policy

Use only calculators on the HKEAA Approved List, bearing the 'H.K.E.A.A. APPROVED' (or older 'H.K.E.A. APPROVED') label. Programmable scientific models (e.g. Casio fx-50FH II, fx-3650P II) are allowed, and you MAY keep your own formulas/programs stored in memory — HKDSE does not require you to clear it. Graphic-display (graphing) and CAS/symbolic calculators are not on the approved list and must not be used.

AO1: Mathematical Knowledge and Understanding (60%)
AO2: Application and Communication (40%)

Built from real past papers and marking schemes (2021–2023).

Tips & Strategies

The 1.5-Minute Rule: Mastering the M1 Clock

In HKDSE Mathematics Module 1, your enemy isn't just the difficulty of the questions—it is the relentless clock. With 150 minutes to complete a 100-mark paper, you have precisely 1.5 minutes per mark. However, top scorers do not allocate their time evenly. They aim to finish the 50 marks of Section A in 60 minutes, leaving a massive 75-minute block for the complex, multi-stage problems in Section B, and a crucial 15-minute buffer at the end to check for arithmetic and rounding slips.

When the exam starts, do not immediately write. Spend 2 minutes scanning Section B. Identify whether the calculus long question features integration by substitution or optimization with curve sketching, and note if the statistics section couples a Poisson rate change with the Central Limit Theorem (CLT). Tackling your strongest topics in Section B first builds momentum and secures high-value marks before cognitive fatigue sets in.

Where the Marks Really Hide: Decimals and Notation

According to yearly HKEAA examiner reports, hundreds of candidates lose easy marks not because of conceptual failure, but due to mathematical hygiene. The most punishing trap is premature rounding. M1 guidelines state that numerical answers must be exact or given to 4 decimal places unless specified otherwise. If you round your intermediate probability values to 2 or 3 decimal places during multi-stage tree diagrams or Bayes' Theorem calculations, your final answer will inevitably drift, throwing away the final accuracy mark (A-mark). Always store intermediate values in your calculator's memory registers (A, B, C, D) or write them down to at least 6 decimal places.

In Calculus, marking schemes are brutally strict regarding notation. Missing the differential \( dx \) or \( dt \) inside integrals, or failing to write down the integration constant \( + C \) in indefinite integration, results in an immediate loss of communication or method marks. When performing integration by substitution, you must explicitly state the differential relationship (e.g., \( du = g'(x)dx \)) and show the updated limits of integration if it is a definite integral. Skipping these steps turns a potential 5-mark subsection into a 1-mark disappointment.

The Concavity Trap: Proving Trapezoidal Estimations

A recurring high-tier question type asks whether the Trapezoidal Rule overestimates or underestimates the true value of a definite integral. Many candidates lose this mark by simply guessing "overestimate" based on a rough mental sketch. To secure full marks, you must rigorously justify your claim using the second derivative \( f''(x) \):

If \( f''(x) < 0 \) (the curve is concave downward) on the interval, the trapezoids lie entirely below the curve, meaning the Trapezoidal Rule will underestimate the true area.
If \( f''(x) > 0 \) (the curve is concave upward), the trapezoids lie above the curve, resulting in an overestimate.

Simply stating the concavity is insufficient. You must explicitly evaluate or show the algebraic sign of \( f''(x) \) across the specified domain. If the question involves a negative coefficient or asks about an inequality, handle the negative signs with extreme care, as they flip the direction of the estimation inequality.

What Top Scorers Do Differently

High achievers treat Statistics not as separate formula blocks, but as a unified system of distributions. When faced with complex conditional probability questions, they immediately sketch a clean probability tree diagram or define events explicitly using capital letters (e.g., let \( A \) be "test is positive", let \( B \) be "has disease"). This guarantees they construct the correct denominator for Bayes' Theorem. Furthermore, they never confuse the Central Limit Theorem standardisation. When standardising a sample mean \( \bar{X} \), the standard error is \( \sigma / \sqrt{n} \), whereas for an individual random variable \( X \), it is simply \( \sigma \). Forgetting to divide the variance by the sample size \( n \) is one of the most common high-frequency errors noted by markers.

Finally, practice utilizing your approved programmable calculator. Having a reliable Trapezoidal Rule program allows you to instantly verify your manual calculations in Section A, ensuring zero arithmetic slips before you turn the page.

Calculator Programs

Binomial Probability

Casio fx-50FH II / fx-3650P II (HKEAA-approved programmable)

Purpose: \(P(X=r)=\binom{n}{r}p^r(1-p)^{n-r}\).

When to use it: Binomial questions asking for an exact number of successes.

Steps

Prompt n, r, p; outputs the probability.

Program

?→N:?→R:?→P:(N nCr R)×P^R×(1-P)^(N-R)

Exam note: p must be between 0 and 1; use nCr (built-in) for the coefficient.

Poisson Probability

Casio fx-50FH II / fx-3650P II (HKEAA-approved programmable)

Purpose: \(P(X=r)=\dfrac{e^{-\lambda}\lambda^r}{r!}\).

When to use it: Rare-event counts over a fixed interval.

Steps

Prompt \(\lambda\), r; outputs the probability.

Program

?→L:?→R:e^(-L)×L^R÷R!

Exam note: \(\lambda\) is the mean; r! uses the factorial key.

Standard Score (z)

Casio fx-50FH II / fx-3650P II (HKEAA-approved programmable)

Purpose: \(z=\dfrac{x-\mu}{\sigma}\) for the standard normal distribution.

When to use it: Before reading the normal distribution table.

Steps

Prompt x, \(\mu\), \(\sigma\); outputs z.

Program

?→X:?→M:?→S:(X-M)÷S

Exam note: \(\sigma\) is the standard deviation, not the variance.

Common Mistakes

1highMarks at stake: 1Standardisation of a normal variable and use of the standard normal table
Truncating intermediate probability values, resulting in final answers that deviate from the required 4 d.p. standard.
How to avoid it: Store intermediate calculation results in the calculator's memory registers (A, B, C, etc.) or write them down to at least 6 decimal places.
2highMarks at stake: 2Sampling distribution and point estimates
Neglecting to divide the variance or standard deviation by \( \sqrt{n} \) when standardising the sample mean under the Central Limit Theorem.
How to avoid it: Always check whether the variable is an individual observation \( X \) or a sample mean \( \bar{X} \). For the latter, use \( \text{Var}(\bar{X}) = \sigma^2/n \).
3mediumMarks at stake: 2Definite integration and its applications
Failing to write down the differential symbol dx or dt, or failing to change the limits of integration when performing integration by substitution.
How to avoid it: Always write the differential step explicitly (e.g., \( du = g'(x)dx \)) and construct a small table to convert \( x \)-limits to \( u \)-limits before integrating.
4mediumMarks at stake: 2Confidence interval for a population mean
Using the sample standard deviation \( s \) directly as the margin of error without dividing by \( \sqrt{n} \) when calculating Confidence Intervals.
How to avoid it: Recall the confidence interval formula: \( \bar{x} \pm z_{\alpha/2} \left(\frac{s}{\sqrt{n}}\right) \). The term \( s/\sqrt{n} \) is mandatory.
5highMarks at stake: 3Conditional probability and Bayes’ theorem
Incorrectly defining the conditional probability space in Bayes' Theorem, especially missing complementary events in the denominator.
How to avoid it: Draw a probability tree diagram immediately. Ensure that all branches leading to the target event are summed in the denominator.

Turn these tips into top grades

thinka turns your weak spots into targeted practice, with instant marking and exam-style feedback. Study smarter, not longer.

Start Practising Free See pricing

Practise real exam questions with instant AI feedback and marking.

Start Practising Free