Algebra, number and structure

VCE Specialist Mathematics Units 3 & 4 — the definitive reference for a 45+ candidate.

Compiled 15 September 2026. Companion to research/sm/01-study-design.md, which this document does not contradict; where that document establishes scope, this one establishes what the examiners actually do with it.

Evidence tags (same convention as 01-study-design.md)

Corpus defects that bear on this document. The 2011 November papers extract with a CID-shifted font encoding and are unreadable as text — their question wordings below are reconstructed from [RPT11] and flagged. The 2024 November paper text files are empty; the 2024 wordings below were read directly from the page images in corpus/sm/pages/. NHT sittings publish answers but no percentages, so the six NHT parts in [QJSON] carry pct = null and can never appear as separators. 2022 Exam 2 Section A Q4 carries pct = null because, per [RPT22 E2], "This question has been redacted following the findings of the Independent Review into the VCAA's Examination-Setting Policies, Processes and Procedures for the VCE."

Contents

---

1. What the study design puts in this area

1.1 The Units 3 and 4 content, verbatim

[SD], overview:

"In this area of study students cover the algebra of complex numbers, including polar form, factorisation of polynomial functions over the complex field and an informal treatment of the fundamental theorem of algebra."

[SD], content (equations linearised from the source OMML):

This area of study includes:
- De Moivre's theorem, proof for integral powers, powers and roots of complex numbers in polar form, and their geometric representation and interpretation
- the nth roots of unity and other complex numbers and their location in the complex plane
- factors over C, of polynomials; and introduction to the fundamental theorem of algebra, including its application to factorisation of polynomial functions of a single variable over C, for example, z⁸ + 1, z² − i or z³ − (2 − i)z² + z − 2 + i
- solution over C of polynomial equations by completing the square, use of the quadratic factorisation and the conjugate root theorem.

That is four dot points, and it is a trap to read them as the whole syllabus. The 2023 design deliberately pushed the elementary machinery down into Units 1 and 2, which [SD] then declares assumed: "Specialist Mathematics Units 3 and 4 assumes familiarity with … the key knowledge and key skills from Specialist Mathematics Units 1 and 2." The Units 1 and 2 area of study with the same name reads, verbatim [SD]:

In this area of study students cover the arithmetic and algebra of complex numbers, including polar form, regions and curves in the complex plane and introduction to factorisation of quadratic functions over the complex field.

Complex numbers. This topic includes:
- definition and properties of the complex numbers, C, arithmetic, modulus of a complex number, and the representation of complex numbers on an Argand diagram
- general solution of quadratic equations (with real coefficients) of a single variable over C, and conjugate roots
- lines, rays, circles and ellipses
- regions defined in the complex plane using combinations of the above
- use of the modulus of a complex number and the argument of a non-zero complex number to prove basic identities
- conversion between Cartesian and polar form of complex numbers
- multiplication, division, and powers of complex numbers in polar form and their geometric interpretation.

Every one of those seven lines is examined in Units 3 and 4. Cartesian arithmetic, conjugates, modulus, argument, conversion, lines, rays, circles, regions — none of them appear in the Units 3 and 4 content list, and all of them appear in the Units 3 and 4 examinations, every year. A student who studies only the four Units 3 and 4 dot points is studying perhaps a third of what is set.

1.2 The Outcome 1 hooks

[SD] attaches key knowledge and key skills to the outcomes, not to the areas of study (see 01-study-design.md §1.1). Three of them are complex-number-specific and they are the formal warrant for most of what the papers do:

Key knowledge:

• complex numbers, Cartesian and polar forms, operations and properties and representation in the complex plane
• geometric interpretation of vectors in the plane and of complex numbers in the complex plane

Key skills:

• perform operations on complex numbers expressed in Cartesian form or polar form and interpret them geometrically
• represent regions of an Argand diagram using complex relations

The bolded key skill is load-bearing. It is the only place in the Units 3 and 4 documentation where loci and regions appear, and it is worth roughly a third of the tagged marks in this area. [OLDSD] had a matching but narrower key skill — "represent curves on an argand diagram using complex relations" — and a content dot point "use of an argand diagram to represent points, lines, rays and circles in the complex plane". The 2023 wording says regions, which is broader, not narrower.

There is also a hook into Area of Study 1 (proof): [SD] names "De Moivre's theorem, proof for integral powers". VCAA has examined exactly that — 2024 NHT Exam 2 Section B Question 2a [PAPERS]:

"Given (cos(θ) + i sin(θ))ⁿ = cos(nθ) + i sin(nθ) for n ∈ Z⁺, show, using appropriate trigonometric identities, that (cos(θ) + i sin(θ))ⁿ⁺¹ = cos((n + 1)θ) + i sin((n + 1)θ) for n ∈ Z⁺"

— which is the inductive step of De Moivre, set as a "show that".

1.3 The formula sheet block, verbatim

[FS] — the whole of what you are given:

Algebra, number and structure (complex numbers)
z = x + iy = r(cos(θ) + i sin(θ)) = r cis(θ)
|z| = √(x² + y²) = r
−π < Arg(z) ≤ π
z₁z₂ = r₁r₂ cis(θ₁ + θ₂)
z₁/z₂ = (r₁/r₂) cis(θ₁ − θ₂)
de Moivre's theorem: zⁿ = rⁿ cis(nθ)

What is not on the sheet, and therefore must be known cold: z z̄ = |z|²; z + z̄ = 2Re(z); z − z̄ = 2i Im(z); |z₁z₂| = |z₁||z₂|; the formula for the nth roots (r^(1/n) cis((θ + 2kπ)/n)); the conjugate root theorem; the area of a circular segment (½r²(θ − sin θ)) — that last one sits in the Mensuration block of [FS], not the complex block, and is needed by roughly one Section B complex question in three.

Note also the formula sheet's convention: cis, never e^{iθ}. Euler's form is nowhere in [SD] or [FS]. It is not wrong to use it in working, but every VCAA answer is expressed in cis.

1.4 What the exams actually test — the counts

[QJSON] contains 235 graded question parts tagged to this area across 2006–2025, worth 374 marks: 78 multiple-choice items and 157 written parts. Distribution:

Marks and share of paper, by sitting:

Four structural regularities, all of them exploitable:

1. Complex numbers has appeared on every single November Examination 1 from 2006 to 2025 — twenty out of twenty — for between 3 and 9 marks, median 4. It is the most reliably scheduled topic in the subject. (The 2026 NHT Examination 1 continues the run: Question 10, 4 marks, "Consider the family of complex polynomials P(z) = z⁴ + bz³ + cz² + dz + 1, where b, c, d ∈ R. Given that P(i) = 0, and that P(z) = 0 has three distinct solutions, find all combinations of the values of b, c and d." [PAPERS])
2. Examination 2 Section B has carried a dedicated complex-numbers extended-response question every year 2006–2025, and in thirteen of the last fourteen it has been Question 2. The exception is 2017, where it was Question 4. Before 2012 it moved around (Q5 in 2006, 2008, 2010; Q1 in 2007, 2011; Q2 in 2009). Worth 9–13 marks.
3. Examination 2 Section A carried 3–5 complex-number items per paper under the five-option format (2006–2022), clustered in a contiguous block — usually Q4–Q9. Under the 2023 four-option format it has settled at exactly two per paper, again adjacent (2023: Q4, Q5; 2024: Q5, Q6; 2025: Q5, Q6).
4. Partial fractions is tagged into this area and floats between the papers. [OLDSD] put "expression of rational functions of low degree as sums of partial fractions" inside the Algebra area of study; the 2023 design moved it to Functions, relations and graphs and into the integration dot points. Either way it is examined — 2022 Exam 1 Question 4 (4 marks), 2024 Exam 1 Question 3a (1 mark), and as a pure form-recognition MCQ in 2007, 2018 and 2020.

How hard is it? Median full-mark rate across all 228 parts with published percentages is 57%. Split:

Two readings. First, Examination 1 complex-number questions are the hardest slice of the topic — the median part is answered fully by fewer than half the state, because Exam 1 removes CAS from exactly the operations (root-finding, polar conversion, surd simplification) that CAS makes trivial. Second, the topic has become easier since 2023 (median 61% against 55% for 2006–2022), largely because the Section A item count halved and the Section B question has drifted toward routine loci work. That drift is worth knowing but should not be relied on: 2023 Exam 2 Section B Question 2fii sat at 7%.

---

2. The complete catalogue of question types

Thirty-three types. For each: the name, the literal VCAA template quoted from a real paper, what it is really testing, the standard method, at least three archive instances with ref and pct, the typical mark value, and the traps the reports name.

Where a quotation reproduces a mathematical expression that the text extraction garbled, the gap is marked ⟨…⟩ rather than reconstructed.

On references. Every reference in the November columns is a verbatim ref value from [QJSON], and every percentage beside it is VCAA's own published full-mark rate. NHT items are cited in the readable form "2018 NHT Exam 1 Q8a" and carry "— (NHT)" in the percentage column: [QJSON] holds only six NHT parts for this area (2024 Exam 2 Section A Q5/Q6/Q18 (NHT) and 2025 Exam 2 Section A Q3/Q4/Q5 (NHT)), and no NHT sitting publishes percentages, so NHT items are quoted for their wording only and can never be separators. The 2026 NHT papers are in [PAPERS] but not in [QJSON] at all.

---

Type 1 — Cartesian arithmetic and the real/imaginary part

Template (2008 Exam 2 Section A Q6 [PAPERS]):

"If z = (3 + 4i)/(1 + 2i), the imaginary part of z is
A. −2 B. −(2/5)i C. −2/5 D. −2i E. 2"

What it tests. That Im(z) is a real number, not bi. The distractor set is built entirely around that confusion: options B and D are the same arithmetic with an i attached.

Method. Multiply numerator and denominator by the conjugate of the denominator; read off the coefficient of i without the i.

Traps. In 2008, 28% chose B — the answer with a spurious i. 2012 Exam 2 Section A Q8 inverts the test: "If z = a + bi, where both a and b are non-zero real numbers and z ∈ C, which of the following does not represent a real number? A. z + z̄ B. |z| C. z z̄ D. z² − 2abi E. (z − z̄)(z + z̄)" — E is a product of a purely imaginary number and a real number, hence imaginary.

---

Type 2 — Division by realising the denominator, answer in a required form

Template (2007 Exam 1 Question 1, 4 marks [PAPERS]):

"Express (2√3 + 2i)/(1 − √3 i) in polar form."

What it tests. Two competing methods and the discipline to pick one. Either rationalise then convert, or convert both parts to polar and divide.

Method (tech-free). Preferred: convert numerator and denominator separately to r cis θ, then divide moduli and subtract arguments. 2√3 + 2i = 4 cis(π/6); 1 − √3 i = 2 cis(−π/3); quotient = 2 cis(π/2).

Traps. [RPT07 E1] is unusually explicit: "it was clear that several students had not learned the relevant exact values for circular functions… Those who converted the numerator and denominator to polar form before proceeding often had difficulty with the quadrant for the denominator. Students should be encouraged to draw small diagrams to indicate in which quadrant the complex number lies." The report also records the error −π/6 − π/3 = −2π/3 → π/3 propagating to the "correct" answer by luck. 2016 Exam 1 Q6 is the same trick in reverse ("Write (1 − √3 i)⁴/(1 + √3 i) in the form a + bi, where a and b are real constants"); [RPT16 E1]: "It was common for the incorrect argument to be used, usually due to the incorrect quadrant but sometimes due to not knowing exact values. A sketch may have been helpful."

---

Type 3 — Conjugate identities as an abstract test

Template (2009 Exam 2 Section A Q6):

"The distance between the two points z and z̄ in the complex plane is given by
A. 2Re(z) B. 2Im(z) C. 2|z| D. 2Re(z) + 2Im(z) E. 2Arg(z)"

What it tests. Whether z z̄ = |z|², z + z̄ = 2Re(z) and z − z̄ = 2i Im(z) are internalised as identities rather than as computations. None of them is on the formula sheet.

Method. Write z = x + iy, z̄ = x − iy, and evaluate directly.

2020 Exam 2 Section A Q5 is the hardest of these on its face: "Given the complex number z = a + bi, where a ∈ R\{0} and b ∈ R, 4z z̄ / (z + z̄)² is equivalent to …" — the answer is 1 + (Im(z)/Re(z))², from 4|z|²/(2Re z)². 2021 NHT Exam 2 Section A Q6 asks for cos(arg(z)) and expects Re(z)/|z|.

Traps. Treating (z + z̄)² as z² + z̄²; forgetting |z|² = z z̄ and expanding into a four-term mess.

---

Type 4 — Modulus of a power, product or quotient; surd modulus

Template (2008 Exam 1 Question 10a, 1 mark [PAPERS]):

"Let w = 1 + ai where a is a real constant.
a. Show that |w³| = (1 + a²)^{3/2}."

What it tests. |wⁿ| = |w|ⁿ. Nothing else.

Method. |w| = √(1 + a²), so |w³| = |w|³ = (1 + a²)^{3/2}.

Traps. [RPT08 E1], on a one-mark "show that" answered by 41%: "It was surprising how many students did not know, or failed to recognise that |w³| = |w|³. Few students who attempted to expand w³ were able to find success." Part b then asked for the values of a with |w³| = 8, and [RPT08 E1] records: "far too many students were unable to simplify expressions involving indices. Of those who correctly obtained a² = 3, several carelessly gave the answer as √3 or −√3" — i.e. one root instead of both. 2026 NHT Exam 2 Section A Q5 is the modern geometric version: "Consider the circle given by |z − 2 − i| = 3, where z ∈ C. The maximum value of |z| for those points that lie on this circle is …" — answer 3 + √5, the distance from the origin to the centre plus the radius.

---

Type 5 — Convert Cartesian to polar form

Template (2006 Exam 1 Question 9a, 1 mark [PAPERS]):

"Express 1 + √3 i in polar form."

and, as a "show that" (2019 Exam 1 Question 7a, 1 mark [PAPERS]):

"Show that 3 − √3 i = 2√3 cis(−π/6)."

What it tests. r = √(x² + y²), θ = the angle in the correct quadrant, expressed as a principal value.

Method (tech-free). Compute r. Compute the reference angle from tan⁻¹(|y|/|x|). Place it in the quadrant by sketching the point. Never trust tan⁻¹(y/x) alone — it cannot distinguish quadrants 1/3 or 2/4.

Traps. [RPT06 E1]: "some had not learned the exact values for trigonometry as the most common mistake was to use π/6 as the argument" (instead of π/3). [RPT07 E2] on Q1a: "The most common error was arg(z₁) = −π/6" for z₁ = −√3 + i, whose argument is 5π/6. [RPT17 E2] on Q4a: "A number of incorrect answers had arguments outside the third quadrant, which should have alerted students to an error, given the signs of the real and imaginary parts. A diagram could have reminded students that the answer needed to be in the third quadrant." [RPT18 E1] on 2018 Exam 1 Q2b: "Quite a few students were not able to write √3 − i in polar form correctly with arguments of π/6, 5π/6 and π/3 being given frequently. Students are reminded that a diagram placing the complex number in the correct quadrant can be helpful in avoiding errors."

---

Type 6 — Convert polar to Cartesian form (exact values)

Template (2009 Exam 1 Question 4, 4 marks [PAPERS]):

"Given that cos(2θ) = 3/4 where θ ∈ ⟨−3π/4, −π/4⟩, find cis(θ) in cartesian form."

What it tests. The double-angle identities as a route to an exact cos θ and sin θ when θ is not a standard angle, plus the quadrant decision on the square root.

Method. Use cos(2θ) = 2cos²θ − 1 to get cos θ = ±√(7)/(2√2); use cos(2θ) = 1 − 2sin²θ (or Pythagoras) for sin θ; choose signs by the given interval for θ.

Traps. [RPT09 E1]: "A large proportion of students did not consider the quadrant of the angle so their final answer contained a sign error. Some students tried to construct a right-angled triangle with the angle 2θ … no progress was made." [RPT06 E2] on the related 2006 Exam 2 Section B Q5b ("Use a double angle formula to show that the exact value of cos(π/8) = √(2 + √2)/2. Explain why any values are rejected"): "Many did not explain adequately why the negative solution to their quadratic in cos(π/8) was rejected." Part c adds the "hence": "A large number of students ignored the 'hence' requirement of this question and did not use the answer for cos(π/8). This was another 'show that' question where the correct steps in manipulating surds needed to be seen."

---

Type 7 — Find the principal argument

Template (2014 Exam 2 Section A Q8):

"The principal argument of (−3√2 − i√6)/(2 + 2i) is
A. −13π/12 B. 7π/12 C. 11π/12 D. 13π/12 E. −11π/12"

What it tests. Two things at once: argument arithmetic under division, and the reduction of the result into (−π, π].

Method. Arg(z₁/z₂) = Arg z₁ − Arg z₂, then add or subtract 2π until the value lies in (−π, π].

Traps. This is the single most persistent source of lost marks in the whole area. [RPT11 E1] on a question whose answer is Arg(z) = 11π/12: "The majority of students attempted to answer the question in cartesian form (using conjugates) rather than converting to polar form, leaving them with little chance of finding the argument… Of those students who correctly got z = 2cis(13π/12), some left this as their answer or gave the principal argument as 13π/12 or −π/12." [RPT09 E2] on Q2c: "A common answer was 7π/6, but the question required the principal value Arg(z)." [RPT25 E2] on Q2d: "Many students stated the correct value for z₀ but did not correctly identify the argument. Students are reminded that drawing the ray on the graph may have made it easier to identify the angle. Principal values were expected." [RPT16 E2] on 2016 Exam 2 Section B Q2f: "Some students did not note that the principal value of the argument was used in the question."

---

Type 8 — Argument arithmetic under powers: Arg(zⁿ)

Template (2008 Exam 2 Section A Q5):

"For a certain complex number z where Arg(z) = π/5, Arg(z⁷) is
A. −7π/5 B. −3π/5 C. 2π/5 D. 3π/5 E. 7π/5"

What it tests. That arg(zⁿ) = n·arg(z) but Arg(zⁿ) ≠ n·Arg(z) in general — the answer is 7π/5 − 2π = −3π/5.

Method. Multiply, then reduce modulo 2π into (−π, π].

Traps. In 2008, 26% chose E — 7π/5, the unreduced value. 2019 Exam 2 Section A Q6 composes it: "Let z, w ∈ C, where Arg(z) = π/2 and Arg(w) = π/4. The value of Arg(z⁵/w⁴) is …" [RPT14 E2] on Q2aii: "many students did not express their answer as an angle in the interval (−π, π]."

---

Type 9 — Multiplication and division in polar form; geometric interpretation

Template (2012 Exam 2 Section A Q5):

"If z = 2cis(4π/5) and w = z⁹, then
A. w = 16√2 cis(36π/5) … E. w = 9√2 cis(4π/5)"

and the geometric version (2015 Exam 2 Section A Q9):

"Let z₁ = r₁cis(θ₁) and z₂ = r₂cis(θ₂), where z₁ and z₁z₂ are shown in the Argand diagram below; θ₁ and θ₂ are acute angles. A statement that is necessarily true is
A. r₂ > 1 B. θ₁ < θ₂ C. |z₁/z₂| > r₁ D. θ₁ = θ₂ E. r₁ > 1"

What it tests. That multiplying by z₂ rotates by θ₂ and scales by r₂ — and that you can read both off a picture.

Method. Moduli multiply, arguments add. To read a diagram: if z₁z₂ is further from the origin than z₁, then r₂ > 1; the rotation angle from z₁ to z₁z₂ is θ₂.

Traps. 2010 Exam 2 Section B Q5c–d is the hardest pair in the archive on this type. Q5c: "Let the argument of u be given by Arg(u) = α. (You are not required to find α.) By expressing iw in polar form in terms of α, show that u/(iw) = 2cis(2α + π)" — 18%. Q5d: "Use the relation given in part a. to find Arg(u + w) in terms of α" — 13%. [RPT10 E2]: "Some students who did attempt the question used approaches such as Arg(u + w) = Arg(u) + Arg(w), which were incorrect." That is the canonical error: arguments add under multiplication, never under addition. The report also notes "Quite a few students did a substantial amount of work in cartesian form to little avail. … As this was a 'show that' question, it was important that all connecting steps were shown."

---

Type 10 — De Moivre: evaluate a power, answer in a + bi

Template (2018 Exam 1 Question 2, 4 marks [PAPERS]):

"a. Show that 1 + i = √2 cis(π/4). (1 mark)
b. Evaluate (√3 − i)¹⁰ / (1 + i)¹², giving your answer in the form a + bi, where a, b ∈ R. (3 marks)"

What it tests. The conversion, the power rule, the reduction of the argument, and the conversion back — four steps, no calculator.

Method. Convert both to polar; apply zⁿ = rⁿ cis(nθ); divide; reduce the argument into (−π, π]; convert back using exact values.

Traps. [RPT18 E1]: "Of those students who obtained the result 16cis(2π/3), some neglected to write the final answer in the required form or made errors in their attempt. A small number of students attempted to expand brackets. This approach was rarely successful." [RPT19 E1] on Q7b: "The efficient method was to use de Moivre's theorem although some students attempted to expand (3 − √3 i)³. Students who chose the latter approach generally did not score as well." Note the 35-point gap between 2018 Exam 1 Q2b (37%, a quotient of two tenth/twelfth powers) and 2019 Exam 1 Q7b (72%, a single cube): the difficulty scales with the number of conversions, not the concept.

---

Type 11 — Find n such that zⁿ is real or purely imaginary

Template (2019 Exam 1 Question 7c–d, 1 mark each [PAPERS]):

"c. Find the integer values of n for which (3 − √3 i)ⁿ is real.
d. Find the integer values of n for which (3 − √3 i)ⁿ = ai, where a is a real number."

What it tests. That "real" means arg = kπ and "purely imaginary" means arg = π/2 + kπ, and that the answer is a general solution with k ∈ Z stated.

Method. With z = r cis θ, zⁿ = rⁿ cis(nθ). Real ⟺ nθ = kπ; purely imaginary ⟺ nθ = (2k + 1)π/2. Solve for n and state the parameter's domain.

Traps. This type is a reliable separator and the reports say why. [RPT19 E1] on 7c: "Some students realised that if n was a positive or negative multiple of 6 then zⁿ was real, but were unable to express this mathematically. Some students did not indicate that k was a member of Z, the set of integers." On 7d: "This question was answered poorly… many students were unable to find a general solution." [RPT06 E2]: "Some realised that the imaginary part of the expression would have to be zero, and others managed to find a few values by trial and error. A large number of students who did manage to do something with this question found only those n values which were positive integers or zero." [RPT12 E2]: "The general solution for n eluded most students. Many gave some specific values for n… Others gave the correct general solution, but failed to define k."

2021 Exam 2 Section A Q6 is the pure form: "If z ∈ C, z ≠ 0 and z² ∈ R, then the possible values of arg(z) are … A. kπ/2, k ∈ Z" — 23%, with the largest distractor C ((2k+1)π/2) at 30%. [RPT21 E2] comments only: "The square of any z with the argument given in option A will be real."

---

Type 12 — De Moivre as a proof device and as a route to trigonometric identities

Template (2024 NHT Exam 2 Section B Q2a, 2 marks [PAPERS]):

"Given (cos(θ) + i sin(θ))ⁿ = cos(nθ) + i sin(nθ) for n ∈ Z⁺, show, using appropriate trigonometric identities, that (cos(θ) + i sin(θ))ⁿ⁺¹ = cos((n + 1)θ) + i sin((n + 1)θ) for n ∈ Z⁺"

and (2023 Exam 2 Section B Q2fii, 2 marks [PAPERS]):

"Given that w = cis(2π/7) satisfies (z − 1)(z⁶ + z⁵ + z⁴ + z³ + z² + z + 1) = 0, use De Moivre's theorem to show that cos(2π/7) + cos(4π/7) + cos(6π/7) = −1/2."

What it tests. The link [SD] explicitly builds between this area and Area of Study 1 ("De Moivre's theorem, proof for integral powers") — and, in the 2023 case, the fact that the sum of the seventh roots of unity is zero.

Method. For the proof: multiply by one more factor of cis θ and expand using the compound-angle formulas. For the identity: the roots of z⁷ = 1 sum to 0 (coefficient of z⁶ is zero), so 1 + Σ cis(2kπ/7) = 0; pair conjugate roots, each pair contributing 2cos(2kπ/7); hence 1 + 2(cos(2π/7) + cos(4π/7) + cos(6π/7)) = 0.

Traps. [RPT23 E2] on Q2fii: "This question was not well done. Many students were able to express the given equation in terms of powers of w but most students did not 'show that' the required result arose through a series of logical steps." The companion part 2023 Exam 2 Section B Q2e ("Verify that the equation z⁷ − 1 = 0 can be expressed in the form (z − 1)(z⁶ + z⁵ + z⁴ + z³ + z² + z + 1) = 0") was answered by 54%, and [RPT23 E2] notes: "Most students expanded the brackets… A significant proportion of students attempted polynomial long division. While some were successful, many did not see the process through to completion."

---

Type 13 — nth roots of a complex number, answers in polar form

Template (2020 Exam 1 Question 3, 3 marks [PAPERS]):

"Find the cube roots of 1/2 − 1/2 i. Express your answers in polar form using principal values of the argument."

What it tests. z^{1/n} = r^{1/n} cis((θ + 2kπ)/n), k = 0, 1, …, n − 1, and then the reduction of each of the n arguments into (−π, π].

Method (tech-free). Convert to polar. Take the real nth root of the modulus. Divide the argument by n to get the first root, then step by 2π/n around the circle. Reduce every argument into (−π, π].

Traps. [RPT20 E1]: "Students should be able to express 1/2 − 1/2 i in polar form by recognition (possibly with the aid of a small diagram). Some students had difficulty with this first step and gave an incorrect argument or modulus. … Some students neglected to give the arguments for their final answers using principal values as required by the question. Some students found the cube of 1/2 − 1/2 i rather than the cube roots." [RPT10 E2] on Q5b: "a number of students misinterpreted the question and found the cube of z₁. Some simplified 200^{1/6} cis(13π/12) to 200^{1/6} cis(−π/12)" — wrong, 13π/12 − 2π = −11π/12 — "and others did not fully apply De Moivre's theorem to the modulus and argument of z₁."

2017 Exam 2 Section A Q4 is the abstract form: "The solutions to zⁿ = 1 + i, n ∈ Z⁺ are given by … E. 2^{1/(2n)} cis(π/(4n) + 2πk/n), k ∈ Z" — the distractors differ only in the exponent on 2 and in whether k ∈ R or k ∈ Z.

---

Type 14 — nth roots in Cartesian form

Template (2015 Exam 1 Question 4a, 3 marks [PAPERS]):

"Find all solutions of z³ = 8i, z ∈ C in cartesian form."

What it tests. The same root machinery, plus exact-value conversion back, plus the discipline to give all roots.

Method. 8i = 8 cis(π/2); roots are 2 cis(π/6), 2 cis(5π/6), 2 cis(−π/2) = √3 + i, −√3 + i, −2i.

Traps. [RPT15 E1] is a checklist of everything that can go wrong: "This question was quite well answered by students who used polar form, but not by the small number of students who tried to solve the equation in cartesian form. … Some students who found the correct solutions in polar form either left them in polar form or converted them to cartesian form with arithmetical errors. Many students assumed that the Conjugate Root Theorem applied. Others tried to use the formula for perfect cubes. Some gave factors rather than solutions, and a number of students gave only one solution for this cubic."

The conjugate-root error is the important one: z³ − 8i = 0 has a non-real coefficient, so conjugate pairs are not forced — and indeed the roots √3 + i, −√3 + i, −2i contain no conjugate pair.

---

Type 15 — Roots of unity: list, plot, and the geometric arrangement

Template (2023 Exam 2 Section B Question 2a–c, 4 marks [PAPERS]):

"Let w = cis(2π/7).
a. Verify that w is a root of z⁷ − 1 = 0. (1 mark)
b. List the other roots of z⁷ − 1 = 0 in polar form. (1 mark)
c. On the Argand diagram below, plot and label the points that represent all the roots of z⁷ − 1 = 0. (2 marks)"

What it tests. That the nth roots of unity are cis(2kπ/n), equally spaced by 2π/n around the unit circle, always including z = 1, and closed under conjugation.

Method. Write them as cis(2kπ/7) for k = −3, −2, −1, 0, 1, 2, 3 (principal values) or k = 0, …, 6. Plot on the unit circle at equal angular spacing.

Traps. [RPT23 E2] on Q2b: "Omitting z = 1 was a common error. A range of equivalent polar forms were seen and accepted." On Q2c: "The majority of students were aware that the roots of unity are evenly spaced around the unit circle. … Some students failed to recognise that the sectors shown had angles of π/6 and incorrectly estimated the required locations." [RPT06 E2] on "Plot the roots of z⁸ = 1 on the Argand diagram": "Common errors involved placing points on the wrong circle, using rays from the origin to show complex numbers, and not plotting all roots."

The "rays from the origin" error recurs across twenty years — see also [RPT06 E2] on Q5ai and [RPT10 E1] on Q4: "Quite a few students drew a line segment from the origin to the point, instead of drawing a dot at the point or a cross at the required location."

---

Type 16 — Transformed root sets ("hence" roots)

Template (2015 Exam 1 Question 4b, 1 mark [PAPERS]):

"Find all solutions of (z − 2i)³ = 8i, z ∈ C in cartesian form."

(following part a, which asked for the solutions of z³ = 8i).

What it tests. That a substitution u = z − 2i turns the new equation into the old one, so the new root set is the old root set translated by +2i. It is a one-mark question that punishes anyone who restarts.

Method. If w₁, w₂, w₃ solve w³ = 8i, then z = wⱼ + 2i.

Traps. [RPT15 E1]: "Students were expected to recognise that the solutions to Question 4a. needed to be translated two units up, and so add 2i. Several students subtracted 2i from the answers in part a., and a small number tried to solve the equation without using their answer to part a."

---

Type 17 — Recognising the root pattern from a diagram

Template (2007 Exam 2 Section A Q6):

"Which one of the following diagrams could represent the location of the roots of z⁵ + z² − z + c = 0 in the complex plane, where c ∈ R?"

What it tests. Two facts simultaneously: real coefficients force conjugate symmetry about the real axis, and a degree-5 polynomial has 5 roots, so at least one must be real.

Method. Reject any diagram that is not symmetric in the real axis; count the points; check at least one lies on the real axis for odd degree.

2022 NHT Exam 2 Section A Q7 is the roots-of-a-complex-number version: "Which one of the following could represent the solutions to z⁵ = 6 + 30i on an Argand diagram?" — here there is no conjugate symmetry (the right-hand side is not real); the roots are five points equally spaced on a circle of radius |6 + 30i|^{1/5}.

Traps. Applying conjugate symmetry when the coefficients are not real. This is the same error [RPT07 E1] names on 2007 Exam 1 Q2b and [RPT24 E1] names on 2024 Exam 1 Q1b.

---

Type 18 — Conjugate root theorem: find the remaining roots

Template (2014 Exam 1 Question 3, 5 marks [PAPERS]):

"Let f be a function of a complex variable, defined by the rule f(z) = z⁴ − 4z³ + 7z² − 4z + 6.
a. Given that z = i is a solution of f(z) = 0, write down a quadratic factor of f(z). (2 marks)
b. Given that the other quadratic factor of f(z) has the form z² + bz + c, find all solutions of z⁴ − 4z³ + 7z² − 4z + 6 = 0 in cartesian form. (3 marks)"

What it tests. The theorem's conclusion (non-real roots of a real-coefficient polynomial come in conjugate pairs) and its hypothesis (the coefficients must be real).

Method. State the conjugate root. Form the quadratic factor (z − α)(z − ᾱ) = z² − 2Re(α)z + |α|². Find the cofactor by equating coefficients (faster and safer than long division). Solve the cofactor.

Traps. [RPT14 E1]: "Most students identified the need to use the conjugate root theorem but some then gave z² − 1 as their answer" (instead of z² + 1). "Confusion between solutions and factors was often evident. … Some quoted z = ±1 as solutions rather than z = ±i." [RPT25 E1] on Q8c is the most useful methodological comment in the archive: "Students who used comparison of coefficients to find the quadratic factor were generally more successful than those who used long or synthetic division. Students who completed the square rather than using the quadratic formula to solve the quadratic equation were also generally more successful." [RPT17 E1] on Q3 lists: "giving a second solution as −1 − i" — the negative rather than the conjugate — "correctly giving 1 + i as a second solution then multiplying this by the given solution to get 2 and stating 2 as the third solution, which was a correct answer but incorrect reasoning."

---

Type 19 — Conjugate root theorem: determine the real coefficients

Template (2025 Exam 2 Section A Q5):

"The equation z³ + az² + bz + 52 = 0, where a, b ∈ R and z ∈ C, has a solution z = 2 − 3i. The value of ab is
A. −232 B. −64 C. −8 D. 0"

What it tests. The hypothesis a, b ∈ R is the whole question; without it the problem is underdetermined.

Method. Conjugate root 2 + 3i; quadratic factor z² − 4z + 13; product of all three roots = −52 so the real root is −52/13 = −4; expand (z + 4)(z² − 4z + 13) and read off a = 0, b = −3. [RPT25 E2] gives both routes: "Either use substitution and then equate coefficients or use the conjugate root theorem."

Traps. 2021 Exam 2 Section B Question 2 is the hardest live instance [PAPERS]:

"The polynomial p(z) = z³ + αz² + βz + γ, where z ∈ C and α, β, γ ∈ R, can also be written as p(z) = (z − z₁)(z − z₂)(z − z₃), where z₁ ∈ R and z₂, z₃ ∈ C.
a. i. State the relationship between z₂ and z₃. (1 mark)
ii. Determine the values of α, β and γ, given that p(2) = −13, |z₂ + z₃| = 0 and |z₂ − z₃| = 6. (3 marks)"

— part i at 73%, part ii at 23%. [RPT21 E2]: "Many students used p(z) in the expanded form, which was less productive than using the factorised form directly. An alternative solution involving purely real z values was possible."

2008 Exam 1 Question 10c is the archive's low-water mark for this type at 3%: "Let p(z) = z³ + bz² + cz + d where b, c and d are non-zero real constants. If p(z) = 0 for z = w and all roots of p(z) = 0 satisfy |z³| = 8, find the values of b, c and d and show that these are the only possible values." [RPT08 E1]: "Some students then found the third root by solving z³ = 8 but the vast majority only considered the solution z = 2, ignoring the solution z = −2. … Very few students realised that z + 2 was also a possibility, but when used, led to a cubic with some zero coefficients (it gives the sum of cubes), which then had to be excluded to fully answer the question." The word "non-zero" in the stem was the entire exclusion argument, and almost nobody used it.

---

Type 20 — Minimum degree, number of distinct roots, and when the theorem does not apply

Template (2009 Exam 2 Section A Q7):

"The polynomial equation P(z) = 0 has real coefficients, and has roots which include z = −2 + i and z = 2. The minimum degree of P(z) would be
A. 1 B. 2 C. 3 D. 4 E. 5"

and its deliberate inverse (2010 Exam 2 Section A Q8):

"The polynomial equation P(z) = 0 has one complex coefficient. Three of the roots of this equation are z = 3 + i, z = 2 − i and z = 0. The minimum degree of P(z) is …"

What it tests. The hypothesis. With real coefficients, three roots are forced (degree 3); with a non-real coefficient, no conjugates are forced (degree 3 as given, not 6).

Traps. 2017 Exam 2 Section A Q3 — "The number of distinct roots of the equation (z⁴ − 1)(z² + 3iz − 2) = 0, where z ∈ C, is A. 2 B. 3 C. 4 D. 5 E. 6" — 47%, with 35% choosing E (6). The quartic gives ±1, ±i; the quadratic factors as (z + i)(z + 2i) giving −i, −2i; −i is a repeat, so there are 5 distinct roots.

2007 Exam 1 Q2 and 2024 Exam 1 Q1 are the two live examples of the hypothesis failing. 2007 [PAPERS]: "a. Show that √5 − i is a solution of the equation z³ − (√5 − i)z² + 4z − 4√5 + 4i = 0. b. Find all other solutions of the equation." [RPT07 E1] on part b (42%): "Far too many students decided that the complex conjugate √5 + i was another solution despite the coefficients of the cubic polynomial not being real. The most efficient method was to factorise by grouping terms in pairs… Of those who found z² + 4 as the quadratic factor, too many solved this to give 4i or 2i as the other two solutions" (rather than ±2i).

2024 [PAPERS], read from the page images: "Consider the function with rule f(z) = 3z³ + 2iz² + 3z + 2i, where z ∈ C. a. Verify that 3z + 2i is a factor of f(z). (1 mark) b. Hence or otherwise, solve the equation f(z) = 0. Give your answers in Cartesian form. (2 marks)". [RPT24 E1] on part b (48%): "some students neglected to show that they were solving an equation and moved directly from the factorised form of the polynomial to writing down the solutions… With the known root, a small number of students tried inappropriately to apply the conjugate root theorem."

---

Type 21 — Factorise over C into linear factors

Template (2010 Exam 1 Question 1, 3 marks [PAPERS]):

"Consider f(z) = z³ + 9z² + 28z + 20, z ∈ C. Given that f(−1) = 0, factorise f(z) over C."

What it tests. That "factorise over C" means linear factors, and that a real quadratic factor with negative discriminant must be split.

Method. Divide out the known linear factor; complete the square or use the quadratic formula on the quadratic; write the answer as a product of three linear factors with brackets intact.

Traps. [RPT10 E1]: "Errors included assuming both (z + 1) and (z − 1) were factors (a kind of spurious notion of a complex conjugate)… Some students omitted brackets, for example (z + 1)(z² + 8z + 16) − 16 + 20 = (z + 1)(z + 4)² + 4 = …, and could not obtain full marks. Some students were unable to factorise z² + 8z + 20 and seemed not to realise that it was necessary to complete the square or use the quadratic formula." [RPT13 E2] on Q2d ("Express z⁴ + 16 as the product of four linear factors in terms of z"): "A number of students confused factors and roots. Some wrote down the correct factors but not as a product as required."

---

Type 22 — Solve a quadratic over C

Template (2022 Exam 1 Question 1, 3 marks [PAPERS]):

"Consider the equation p(z) = z² + 6iz − 25, z ∈ C.
a. Express p(z) in the form p(z) = (z + ai)² + b, where a, b ∈ R. (1 mark)
b. Hence, or otherwise, find the solutions of the equation p(z) = 0. (2 marks)"

What it tests. Completing the square when the linear coefficient is imaginary — the study design's own listed method ("solution over C of polynomial equations by completing the square").

Method. (z + 3i)² = z² + 6iz − 9, so p(z) = (z + 3i)² − 16; then z = −3i ± 4.

Traps. The "show that" variants are where the marks go. [RPT19 E2] on Q2ai ("Show that the solutions of 2z² + 4z + 5 = 0, where z ∈ C, are z = −1 ± (√6/2)i"): "In a 'show that' question such as this, students are expected to explicitly show that the given information leads to the required conclusion rather than 'verify' that the given values of z are solutions of the equation." [RPT07 E2] on Q1c ("By solving z² − 2√3 z + 4 = 0 algebraically, show that the roots…"): "The major error in this question was verifying the solutions by substitution, which was contrary to the explicit instruction, 'by solving algebraically' given in the question."

2006 Exam 1 Q9b — "Solve the quadratic equation z² + 2z − √3 i = 0, expressing your answers in exact cartesian form" — is the hardest quadratic in the archive at 16%, because the discriminant 4 + 4√3 i is not real and so its square root has to be found via polar form (part a had supplied 1 + √3 i = 2cis(π/3)). [RPT06 E1]: "A large proportion of students stopped after completing the square or using the quadratic formula, making it clear that many did not know that cartesian form means in the form x + iy where x and y are real numbers."

2024 NHT Exam 2 Section A Q5 is the modern Section A version with a non-real linear coefficient: "For z ∈ C and a ∈ R, the discriminant of the quadratic equation az² − aiz − 5 = 0 is 36."

---

Type 23 — Equations reducible to quadratics; square roots of a complex number

Template (2013 Exam 1 Question 8, 4 marks [PAPERS]):

"Find all solutions of z⁴ − 2z² + 4 = 0, z ∈ C in cartesian form."

What it tests. Treat as a quadratic in z², obtain z² = 1 ± √3 i, then take square roots of a non-real number — the study design's own example z² − i.

Method (tech-free). Two routes. (i) Polar: 1 + √3 i = 2cis(π/3), so z = √2 cis(π/6) and √2 cis(−5π/6); convert with exact values. (ii) Cartesian: set (x + iy)² = 1 + √3 i, giving x² − y² = 1 and 2xy = √3 — solvable but algebraically heavy.

Traps. [RPT13 E1]: "The most frequently occurring answer from that point was z = ±√(1 ± √3 i), which is not of the form z = x + iy. Those who used polar form often achieved correct answers, although a few forgot to change back to Cartesian form. Some students made errors in the sine and cosine of standard angles. … Solving (x + iy)² = 1 ± √3 i was a viable approach but most who used this approach struggled with the algebra." [RPT09 E1] on the easier z⁴ − z² − 6 = 0: "quite a few forgot to include ± or the square root signs. Some incorrectly wrote ±2i. A significant number of students attempted to find the correct answer but then discarded some solutions, stating that z² = −2 had no solutions."

2021 Exam 1 Question 8 is a different reduction and the second-hardest written part of the last decade [PAPERS]: "a. Solve z² + 2z + 2 = 0 for z, where z ∈ C. (1 mark) b. Solve z² + 2z̄ + 2 = 0 for z, where z ∈ C. (3 marks)" — 70% then 10%. [RPT21 E1]: "Students who were successful let z = x + iy, leading to two simultaneous real equations. Algebraic errors were often seen… A number of students assumed that the solutions to part a. were also solutions to part b., and some students confused the complex conjugate with the reciprocal."

---

Type 24 — Circles: |z − a| = r and z z̄ = r²

Template (2018 Exam 2 Section B Question 2a, 1 mark [PAPERS]):

"State the centre in the form (x, y), where x, y ∈ R, and state the radius of the circle given by |z − (1 + 2i)| = 2, where z ∈ C."

and (2025 Exam 2 Section B Question 2a, 1 mark [PAPERS]):

"Sketch {z : z z̄ = 4, z ∈ C} on the Argand plane below."

What it tests. Two equivalent presentations of the circle: |z − a| = r (centre a, radius r) and z z̄ = r² ⟺ |z|² = r² (centre origin, radius r, not r²).

Method. Read the centre and radius directly; or substitute z = x + iy and complete the square. For axis intercepts, set x = 0 or y = 0 in the Cartesian form.

Traps. 2010 Exam 2 Section A Q10 — "On an argand diagram, a set of points which lies on a circle of radius 2 centred at the origin is A. {z ∈ C : z z̄ = 2} B. {z ∈ C : z² = 4} C. {z ∈ C : Re(z²) + Im(z²) = 4} D. {z ∈ C : (z + z̄)² − (z − z̄)² = 16} E. {z ∈ C : (Re z)² − (Im z)² = 16}" — 28% correct (D), with 42% choosing C. Option A is the classic trap: z z̄ = 2 is the circle of radius √2, not 2. [RPT18 E2] on Q2a: "Some students gave only one of the two required parts of the answer. An incorrect radius of √2 was occasionally given." [RPT24 E2] on Q2c: "Students should be mindful that the circle should be drawn smoothly through the four extreme points and should not have a pointed shape."

---

Type 25 — The perpendicular bisector |z − a| = |z − b|

Template (2025 Exam 2 Section B Q2bi, 2 marks [PAPERS], read from the PDF — the plain-text extraction drops the surd and renders it as |z − 3 − i|):

"Show that {z : |z − 2i| = |z − √3 − i|, z ∈ C} may be expressed as y = √3 x."

What it tests. That "equidistant from two points" is a straight line — specifically the perpendicular bisector of the segment joining them.

Method. Two routes, and the reports say both are accepted. Algebraic: substitute z = x + iy, square both sides, expand; the x² and y² terms cancel, leaving a linear equation. Geometric: find the midpoint of a and b, take the negative reciprocal of the gradient of ab, write the line.

Traps. [RPT25 E2] gives the full model answer for both routes: "The algebraic approach is to equate the magnitudes of the complex expressions and then expand the brackets and simplify. The geometric approach required finding the midpoint and the gradient of the line segment." [RPT16 E2] on Q2a: "The majority of correct answers resulted from substituting z = x + yi into the expression provided. Very few students used a perpendicular bisector approach at this stage. The most common error was a negative gradient." [RPT09 E2] on Q2b: "a large number of students did not show the full expansion which led to the given result." [RPT15 E2] on Q2aii: "Few students seemed to realise that the required line was the perpendicular bisector of the line interval joining (0, 0) and (1, √3)."

2020 Exam 2 Section B Q2c adds the interpretive part (1 mark): "State a geometrical interpretation of the graph of |z − u| = |z − v| in relation to the points that represent u and v."

---

Type 26 — The Apollonius circle |z − a| = k|z − b|

Template (2018 Exam 2 Section B Question 2b, 2 marks [PAPERS]):

"By expressing the circle given by |z + 1| = 2|z − i| in cartesian form, show that this circle has the same centre and radius as the circle given by |z − (1 + 2i)| = 2."

What it tests. That |z − a| = k|z − b| is a circle when k ≠ 1 and a line when k = 1 — the single most-tested distinction in the loci family.

Method. Substitute z = x + iy, square both sides (the factor k² multiplies one side), collect: the x² and y² terms survive with coefficient 1 − k²; divide through and complete the square.

Traps. [RPT18 E2]: "Most students were able to correctly find an expression that did not involve i. In a 'show that' question such as this, students are expected to explicitly show that the given relation leads to the required conclusion." 2015 Exam 2 Section A Q8 is the discriminating item: "A relation that does not represent a circle in the complex plane is A. z z̄ = 4 B. |z + 3i| = 2|z − i| C. |z − i| = |z + 2| D. |z − 1 + i| = 4 E. z + 2z̄ = 4" — the answer is E (a line), with C also a line but excluded by "a relation" phrasing in the published key; the item sat at 57%.

2014 Exam 2 Section B Q2bii is the (z + 2i)(z̄ − 2i) = 4 disguise: "Show that the relation (z + 2i)(z̄ − 2i) = 4 can be expressed in cartesian form as x² + (y + 2)² = 4." [RPT14 E2]: "Most students could express the relation in terms of x and y, but a large number could not follow through with enough mathematical detail to show the given result. Some students substituted the incorrect forms z̄ = x + y and z = x − y."

---

Type 27 — Rays: Arg(z − z₀) = θ

Template (2024 Exam 2 Section B Question 2d, 2 marks [PAPERS], read from the page image):

"A ray originating at the point z = 2 − i passes through the point z = −2 + 3i, cutting the second circle into two segments.
d. i. Sketch the ray on the Argand diagram provided in part c. (1 mark)
ii. Find the equation of the ray in the form Arg(z − z₀) = θ where z₀ ∈ C and θ is measured in radians in terms of π. (1 mark)"

What it tests. Three things that get separated in the marking: the correct origin z₀, the correct angle as a principal value, and the fact that a ray is a half-line excluding its endpoint.

Method. z₀ is the point of emanation. θ = Arg(z₁ − z₀) for any other point z₁ on the ray. Draw an open circle at z₀. In Cartesian form the rule is y − y₀ = tan(θ)(x − x₀) with a stated domain restriction — x > x₀ or x < x₀ — because half the line is not on the ray.

Traps. This is the densest cluster of separators in the whole area — nine of the eleven instances above are at or below 56%, and five are separators. The reports are consistent:

• Endpoint. [RPT21 E2]: "Where drawn, the ray generally had the correct argument. The point of emanation is not part of the required ray and should be shown as an open circle. This was not always shown or placed correctly."
• Line instead of ray. [RPT16 E2]: "Many students were not able to sketch the required ray. Some students sketched a line but did not restrict their ray appropriately, either including or extending past the origin." [RPT20 E2]: "Incorrect responses frequently extended through the point representing u; in some cases, a line was sketched instead of a ray." [RPT22 E2]: "Many students did not draw a ray; in some cases this appeared to be an unfortunate slip as some of these gave a correct argument."
• Domain omitted. [RPT20 E2] on Q2dii, 25%: "While a high proportion of students gave the correct rule, many did not fully describe the function as they did not include the domain."
• Wrong angle. [RPT24 E2] on Q2dii: "The most common error was to quote the argument as −π/4" (rather than 3π/4). [RPT23 E2] on Q2dii, 18%: "While many students correctly identified z₀, finding the correct angle was a challenge for most."

2022 Exam 2 Section A Q5 is the Cartesian-conversion MCQ: "Let z = x + yi, where x, y ∈ R. If Arg(z − i) = 3π/4, which one of the following is true? A. y = 1 − x, x < 0 … E. y = 1 + x, x < 0" — the point of the item is the domain restriction, not the line.

---

Type 28 — Intersections of two loci

Template (2025 Exam 2 Section B Q2ci, 2 marks [PAPERS]):

"Find the points of intersection of the curves defined in part a and in part b.i, expressing your answers in the form a + ib, where a, b ∈ R."

What it tests. Converting both relations to Cartesian form and solving simultaneously — then converting back to a + ib.

Method. Get both in x, y. Substitute. Solve the resulting quadratic. Give both points. Write each as a + ib, not as an ordered pair, when the question says "in the form a + ib".

Traps. [RPT13 E2] on Q2b (15%) — the intersection of z z̄ = 4 with |z + z̄| = |z − z̄|, which is the pair of lines y = ±x: "Most students found only two solutions for z, z = √2 + i√2 and z = −√2 − i√2" — i.e. they used only y = x and missed y = −x entirely; the answer has four points. [RPT08 E2] on Q5c: "An algebraic attempt at the solution was often seen in this question, but this was followed through successfully in only a small number of cases. A fairly common error was interchanging x and y coordinates in the second point of intersection." [RPT25 E2] on Q2cii (labelling): "Students need to take care when labelling the points on the graph. If they are using coordinates, they must not have i in the coordinate."

2019 Exam 2 Section A Q5 is the two-rays version: "The rays Arg(z − 2) = π/4 and Arg(z − (5 + i)) = 5π/6, where z ∈ C, intersect on the complex plane at a point (a, b). The value of b is …" — 38%, with 25% on B and 21% on E.

---

Type 29 — Regions defined by inequalities, and shading

Template (2013 Exam 2 Section B Question 2e, 1 mark [PAPERS]):

"On the Argand diagram provided in part a., shade the region defined by {z : |z| ≤ 2, z ∈ C} ∩ {z : Re(z) ≥ √2, z ∈ C}"

and ([SAMPLE] Exam 2 Section B Q1c, 1 mark):

"On the Argand diagram in part b., shade the region defined by {z : z z̄ ≤ 4, z ∈ C} ∩ {z : Re(z) + Im(z) ≥ 2, z ∈ C}"

What it tests. Reading intersection notation; knowing which side of each boundary satisfies the inequality; and whether boundaries are included (solid) or excluded (dashed).

Method. Sketch each boundary as an equality. Test a convenient point (usually the origin) in each inequality. Shade only the overlap. Use a solid boundary for ≤/≥, dashed for </>.

Traps. [RPT13 E2]: "Frequent errors included segments shaded in other quadrants, the major segment shaded, shading of an annulus, and inaccurate borders and shading of the defined region." [RPT12 E2]: "poor shading of the required area was often seen and some students omitted the corner points from the region." [RPT07 E2] on Q1g (28%): "the most common error being inaccurate placement of the lower corner point of the region. Many students simply shaded various regions inside a circle of radius 2." [RPT08 E2]: "A pleasing number of students managed to shade the region 0 ≤ Arg(z) ≤ 2π/3 correctly."

2012 Exam 2 Section B Q2c is the annulus-sector template: "On the Argand diagram below, shade the region defined by {z : Arg(z₁) ≤ Arg(z) ≤ Arg(z₁⁴)} ∩ {z : 1 ≤ |z| ≤ 2}, z ∈ C."

---

Type 30 — Areas of regions in the complex plane

Template (2024 Exam 2 Section B Q2e, 2 marks [PAPERS]):

"Find the area of the minor segment formed by the intersection of the ray and the circle."

What it tests. The circular-segment formula A = ½r²(θ − sin θ) from the Mensuration block of [FS], applied with the correct central angle.

Method. Find the two intersection points. Find the central angle θ subtended at the centre (not the angle of the ray, and not the inscribed angle). Apply ½r²(θ − sin θ) for a segment, ½r²θ for a sector, ½(r₂² − r₁²)θ for an annulus sector. Give an exact answer unless told otherwise.

Traps. The recurring failure is the angle. [RPT17 E2] on Q4g (24%): "A significant number of students incorrectly used a sector angle of π/3. Solutions using definite integrals were also seen; these solutions were usually completed correctly." [RPT24 E2]: "Most students were successful when applying the area formula of a segment. Some students used the incorrect angle." [RPT25 E2] repeats it verbatim and adds "others did not include the answer in the required form."

The second failure is reaching for integration. [RPT16 E2] on Q2d: "A correct answer was most easily found by adding a right-angled triangle to three-quarters of a circle. … a larger number set up elaborate definite integrals to find the area, occasionally successfully, but this was not an efficient approach." [RPT09 E2] on Q2f (25%): "Few students realised that the area of a portion of an annulus was to be found. Often elaborate approaches were set up to solve this simple problem."

---

Type 31 — Reverse engineering: identify the relation from a picture, or classify relations

Template (2006 Exam 2 Section A Q6):

"The region represented on the above Argand diagram, where a is a real constant, could be defined by
A. |z − (a + 2i)| ≤ 1 B. |z − (a + 2i)| ≥ 1 C. |z − (−a + 2i)| ≤ 1 D. |z − (a + 2i)| ≤ 2 E. |z + a − 2i| ≤ 1"

and the classification form (2006 Exam 2 Section A Q7):

"Which one of the following relations does not have a graph that is a straight line passing through the origin?
A. z + z̄ = 0 B. 3Re(z) = Im(z) C. |z| = |iz̄| D. Re(z) − 2Im(z) = 0 E. Re(z) + Im(z) = 1"

What it tests. Reading |z − a| = r off a diagram with the correct sign on a; and recognising the standard families (|z| = r circle, |z − a| = |z − b| line, Arg(z − z₀) = θ ray, Re/Im linear relations).

Traps. 2015 Exam 2 Section A Q6 at 43% — "Which one of the following relations has a graph that passes through the point 1 + 2i in the complex plane? A. z z̄ = 5 B. Arg(z) = π/3 C. |z − 1| = |z − 2i| D. Re(z) = 2Im(z) E. z + z̄ = 2" — punishes anyone who tests the point numerically without care: z z̄ = 1 + 4 = 5 ✓. Option D is 1 = 4 ✗ and E is 2 = 2 ✓ for Re(z) = 1 — so the item turns on reading "passes through 1 + 2i", not "Re(z) = 1".

2014 Exam 2 Section A Q9 — "The circle |z − 3 − 2i| = 2 is intersected exactly twice by the line given by A. |z − i| = |z + 1| B. |z − 3 − 2i| = |z − 5| C. |z − 3 − 2i| = |z − 10i| D. Im(z) = 0 E. Re(z) = 5" — options B and C are perpendicular bisectors of a segment with the centre as one endpoint, hence they pass through the interior; D and E are tangent or disjoint.

---

Type 32 — Complex multiplication as a rotation; plotting z, z², iz, z̄, −z̄

Template (2012 Exam 2 Section A Q6):

"For any complex number z, the location on an Argand diagram of the complex number u = i³z̄ can be found by
A. rotating z through 3π/2 in an anticlockwise direction about the origin
B. reflecting z about the x-axis and then reflecting about the y-axis
C. reflecting z about the y-axis and then rotating anticlockwise through π/2 about the origin
D. reflecting z about the x-axis and then rotating anticlockwise through π/2 about the origin
E. rotating z through 3π/2 in a clockwise direction ⟨…⟩"

and the constructive version (2010 Exam 1 Question 4, 3 marks [PAPERS]):

"Given that z = 1 + i, plot and label points for each of the following on the argand diagram below. i. z ii. z² iii. z⁴"

What it tests. Multiplication by cis θ = rotation by θ; multiplication by i = rotation by +π/2; conjugation = reflection in the real axis; −z̄ = reflection in the imaginary axis.

Traps. In 2012 the answer is C and 27% chose D — the two differ only in which axis the reflection uses. [RPT10 E1] on the plotting version: "The most common error was to place z on the unit circle. Other errors were z⁴ being placed at (4, 0) or at (0, −4), z² being placed in two locations and z⁴ being placed in four locations." (z = 1 + i has modulus √2, so z² = 2i and z⁴ = −4.)

2024 Exam 2 Section A Q5 [PAPERS] (read from the page image): "If the point z = 1 + √3 i is represented on an Argand diagram, the point representing −z̄ can be located by A. reflecting the point representing z in the real axis. B. rotating … anticlockwise about the origin by 90°. C. reflecting the point representing z in the imaginary axis. D. rotating … clockwise about the origin by 90°." — 70%.

2018 Exam 2 Section A Q6 is the area version: "The complex numbers z, iz and z + iz, where z ∈ C \ {0}, are plotted in the Argand plane, forming the vertices of a triangle. The area of this triangle is given by …" — the answer is |z|²/2, since z and iz are perpendicular and of equal length.

---

Type 33 — Partial fractions set as algebra

Template (2018 Exam 2 Section A Q3):

"Which one of the following, where A, B, C and D are non-zero real numbers, is the partial fraction form for the expression (2x² + 3x + 1)/((2x + 1)³(x² − 1))?"

and (2022 Exam 1 Question 4, 4 marks [PAPERS]):

"Find ∫ (3x² + 4x − 12)/(x(x² − 4)) dx."

What it tests. The form — a repeated linear factor needs one term per power; an irreducible quadratic needs a linear numerator; a factorisable quadratic should be split first.

Method. Factorise the denominator completely over R. Assign A/(px+q) + B/(px+q)² + … for repeated linear factors, and (Ax + B)/(quadratic) only when the quadratic is irreducible. Then equate coefficients or substitute convenient values.

Traps. 2020 Exam 2 Section A Q7 at 26% — "For non-zero real constants a and b, where b < 0, the expression 1/(ax(x² + b)) in partial fraction form with linear denominators, where A, B and C are real constants, is …" — turns entirely on b < 0, which makes x² + b = (x − √(−b))(x + √(−b)) factorisable over R. [RPT22 E1] on the integral: "A small number of students realised that 3x² + 4x − 12 could be split so as to remove the need to use partial fractions. Many students used elements of both methods… Such approaches were inefficient… A number of students did not include absolute value signs in the logarithmic term or failed to include the arbitrary constant."

---

A note on two parts in [QJSON] that are tagged to this area but are not complex-number questions. 2008 Exam 2 Section B Q1dii (33%) is a change-of-variable in a volume-of-revolution integral, and 2010 Exam 2 Section B Q1di (67%) is a vector midpoint. 2007 Exam 1 Q6b (63%), 2012 Exam 2 Section B Q1a (73%) and 2019 Exam 2 Section B Q1a (76%) are parametric-to-Cartesian conversions, tagged here because the old Algebra area absorbed them. They are listed for completeness and are covered properly in the Calculus and Space-and-measurement references.

---

3. The standard wordings

VCAA reuses a small closed set of sentences. Learning them is worth marks, because each one names a required form and the reports are explicit that answers in the wrong form do not score.

One correction that matters. The phrase "sketch the subset of the complex plane defined by" does not occur anywhere in this corpus. A case-insensitive search for subset across all 97 Specialist Mathematics paper and report files returns zero hits, and the word appears in [SD] only in the transformations and counting topics of Units 1 and 2 ("effect of these linear transformations … on subsets of the plane", "countable and uncountable subsets of R"). VCAA's actual wordings for this task are:

• "On the Argand diagram below, sketch {z : z z̄ = 4, z ∈ C} and sketch {z : |z + z̄| = |z − z̄|, z ∈ C}" (2013 Exam 2 Section B Q2a)
• "Sketch {z : z z̄ = 4, z ∈ C} on the Argand plane below" (2025 Exam 2 Section B Q2a)
• "Sketch the circle on the Argand diagram below" (2018 NHT Exam 1 Q8a)
• "Sketch L and the graph of |z| = 2 on the argand diagram below" (2009 Exam 2 Section B Q2e)
• "Graph the circle given by |z + 1| = 2|z − i| on the Argand diagram below, labelling the intercepts with the vertical axis" (2018 Exam 2 Section B Q2c)

Set-builder notation with z ∈ C is the house style from 2012 onward; before that VCAA wrote "the relation …" or "the curves given by …". Two more house conventions worth internalising: VCAA writes "Argand diagram" in papers up to 2023 and "Argand plane" in 2025 and 2026 NHT, interchangeably; and it lowercases "argand" in several 2006–2010 papers. Neither carries meaning.

---

4. The separators in this area

Definition. A separator is a question part with pct ≤ 50 — fewer than half the state earned full marks. In this area there are 90 of them out of 228 parts with published percentages (39.5%). Nineteen are multiple-choice; seventy-one are written. Every one is listed below, grouped by type, as ref — pct% — description.

4.1 Loci, regions and areas in the complex plane — 33 separators

4.2 Polar form, argument and De Moivre — 28 separators

4.3 Roots of complex numbers — 5 separators

4.4 Polynomials over C — 18 separators

4.5 Partial fractions — 4 separators

4.6 Tagged to this area but not complex-number content — 2 separators

4.7 Which types separate most

Ranking by the proportion of instances that fall at or below 50%, and by depth:

What the reports say went wrong, by frequency of complaint. Counting the diagnoses in the 90 separator comments in [QJSON]:

1. The general solution was not given, or the parameter was not defined — [RPT19 E1]: "some students did not indicate that k was a member of Z"; [RPT12 E2]: "Others gave the correct general solution, but failed to define k"; [RPT06 E2]: "found only those n values which were positive integers or zero."
2. The principal-value reduction was not done — [RPT11 E1], [RPT09 E2], [RPT20 E1], [RPT25 E2], [RPT16 E2], [RPT14 E2] all name it.
3. The wrong quadrant — [RPT07 E1], [RPT16 E1], [RPT17 E2], [RPT18 E1]; every one of them recommends drawing a sketch.
4. "Show that" treated as "verify" — [RPT07 E2], [RPT09 E2], [RPT10 E2], [RPT14 E2], [RPT18 E2], [RPT19 E2], [RPT22 E2], [RPT23 E2]. [RPT22 E2] on Q2ai: "A number of students apparently used a CAS to solve the given equation and then substituted their answers, again using CAS to verify the given result."
5. Solutions confused with factors — [RPT07 E1], [RPT12 E1], [RPT13 E2], [RPT14 E1], [RPT15 E1], [RPT17 E1].
6. Conjugate root theorem applied where the coefficients are not real — [RPT07 E1], [RPT15 E1], [RPT24 E1].
7. Not all solutions given — [RPT13 E2] (two of four), [RPT15 E1] ("only one solution for this cubic"), [RPT08 E1] ("ignoring the solution z = −2"), [RPT19 E2] ("most of these students found only one end point of the interval").
8. Exact form abandoned — [RPT13 E2], [RPT18 E2], [RPT24 E2], [RPT25 E2].
9. Rays drawn as lines, or with the endpoint included — [RPT16 E2], [RPT20 E2], [RPT21 E2], [RPT22 E2].
10. Plotting errors: wrong circle, rays from the origin, unlabelled points — [RPT06 E2], [RPT07 E2], [RPT08 E2], [RPT10 E1], [RPT12 E1], [RPT13 E2], [RPT24 E1].

---

5. What makes a hard one hard

Five mechanisms account for nearly every separator in this area. They are worth studying as mechanisms, because the same five recombine year after year.

5.1 Choosing the argument branch

The formula sheet gives you −π < Arg(z) ≤ π and nothing else. Every difficulty follows.

• Converting to polar. tan⁻¹(y/x) returns a value in (−π/2, π/2) and cannot distinguish the first quadrant from the third, or the second from the fourth. The only reliable procedure is: compute the reference angle from tan⁻¹(|y|/|x|), sketch the point, and place the angle by quadrant. Four separate reports across fourteen years tell students to draw the sketch; the advice is not decorative.
• After a power or a quotient. Arg(zⁿ) is n·Arg(z) reduced into (−π, π]; the reduction is a separate step that is silently skipped. 2011 Exam 1 Q4 (23%) hangs entirely on 13π/12 → −11π/12, and [RPT11 E1] records students giving 13π/12 and −π/12 — the second of which is the correct reduction of −π/12's reference angle, not of 13π/12.
• When the question asks for a general solution. zⁿ real means n·θ = kπ for some integer k, and the answer is a set, not a number. The instruction to state k ∈ Z is worth a mark in itself.
• In a ray equation. Arg(z − z₀) = θ requires θ as a principal value, so a ray pointing down and to the left is −3π/4, never 5π/4. [RPT24 E2] records −π/4 given for a ray whose argument is 3π/4 — a sign error and a quadrant error at once.

The discipline: after every argument computation, write down the interval (−π, π] and check the answer against it, and against the quadrant of the point.

5.2 Exact values off the unit circle

Examination 1 forbids technology, and every Exam 1 report from 2006 to 2017 lists "knowing the exact values for circular functions" among the areas of weakness, naming three to five questions each time. This is not a complex-numbers problem; it is a Year 10 trigonometry problem that complex numbers expose.

The standard angles (π/6, π/4, π/3 and their reflections) are assumed. What separates is the non-standard angle:

• 2009 Exam 1 Q4 (32%) sets cos(2θ) = 3/4, so θ is not a standard angle at all and both cos θ and sin θ must be built from the double-angle identities with the sign chosen by the given interval.
• 2006 Exam 2 Section B Q5b–c (29%, 37%) builds cos(π/8) and sin(π/8) from cos(π/4) — the half-angle case — and then requires a justification for rejecting the negative root.
• 2012 Exam 2 Section B Q2a (73%) runs the same construction in reverse for π/12, and the same question's part e asks for a general n at 25%.

The discipline: a half-angle or double-angle bridge is a standard Specialist move, not an emergency measure. When the argument is π/8, π/12, 3π/8 or 5π/12, expect to construct the exact value rather than recall it — and expect a "show that" instruction, which means the surd manipulation must be visible.

5.3 Open versus closed boundaries when sketching

Three distinct decisions, each separately marked:

1. Ray endpoints are excluded. Arg(z − z₀) = θ is undefined at z = z₀ because Arg(0) does not exist. The point of emanation must be drawn as an open circle, and [RPT21 E2] says so in exactly those words. 2021 Exam 2 Section B Q2b sat at 31% largely because of this.
2. Inequalities: solid or dashed. |z − a| ≤ r includes its boundary circle; |z − a| < r does not. VCAA's region questions have used ≤/≥ almost exclusively, so the boundary is normally solid — but the corner points where two boundaries meet must be shown and, when both inequalities are non-strict, included. [RPT12 E2]: "some students omitted the corner points from the region."
3. Domain restrictions on the Cartesian rule of a ray. 2020 Exam 2 Section B Q2dii asked for "the function that describes the ray Arg(z − u) = π/4, giving the rule in cartesian form" and sat at 25%. [RPT20 E2]: "While a high proportion of students gave the correct rule, many did not fully describe the function as they did not include the domain." The rule y = x + 1 is only half an answer; y = x + 1, x > −2 is the answer.

A fourth, related decision: major versus minor. 2016 Exam 2 Section B Q2d asked for the major segment (39%) and 2017 Exam 2 Section B Q4g for the major segment (24%), while 2024 and 2025 asked for the minor one. [RPT13 E2] lists "the major segment shaded" as a frequent error on a minor-segment question. Read the adjective.

5.4 Keeping conjugate pairs — and knowing when not to

The conjugate root theorem has a hypothesis: the polynomial must have real coefficients. VCAA tests the hypothesis at least as often as the conclusion, and [SD] signals this by choosing z³ − (2 − i)z² + z − 2 + i as one of its three worked factorisation examples — a cubic with non-real coefficients.

Live instances where the theorem does not apply:

And where it does apply, the failure mode is symmetric: writing −α instead of ᾱ. [RPT17 E1] records "giving a second solution as −1 − i" where the correct conjugate of 1 − i is 1 + i. [RPT14 E1] records z² − 1 written where z² + 1 was required.

The discipline: before invoking the theorem, read the coefficients. If any of them is non-real — including a lone i buried in a middle term — the theorem is off the table and you must factorise directly, by grouping, by the factor theorem, or by division.

5.5 Surd algebra inside modulus calculations

The modulus turns every loci problem into a surd problem, and the reports name algebra and arithmetic as areas of weakness on almost every Exam 1. [RPT16 E1], in a list of areas of weakness:

"algebraic skills. Difficulty with algebra was evident in several questions. The inability to simplify expressions often prevented students from completing the question. Incorrect attempts to factorise, expand and simplify were common. Poor use of brackets was also common.
arithmetic skills. Difficulty with arithmetic was evident in several questions. The inability to evaluate expressions, especially those involving fractions or surds, was common."

and [RPT07 E2], listing weaknesses, names "basic skills in the areas of plotting complex numbers, their negatives and conjugates, applying double angled and related formulae and manipulating surds."

The specific pressure points:

• |z − a| = |z − b| expanded. Squaring both sides produces x² − 2ax + a² + y² − 2by + b² on each side; the quadratic terms cancel only if you expand correctly. [RPT09 E2]: "a large number of students did not show the full expansion which led to the given result."
• |z − a| = k|z − b| expanded. Here the quadratic terms do not cancel; they survive with coefficient 1 − k², and the whole equation must be divided by it before completing the square. This is where 2018 Exam 2 Section B Q2b (54%) loses its marks.
• Rationalising a surd modulus. 2010 Exam 2 Section B Q5a asks students to show |z₁| = 10√2 where z₁ = (u + w)u/(iw); [RPT10 E2] records "a large number of students got an incorrect answer for z₁."
• Index laws on a fractional-power modulus. 2008 Exam 1 Q10b (59%): [RPT08 E1] "far too many students were unable to simplify expressions involving indices" — the step (1 + a²)^{3/2} = 8 ⟹ 1 + a² = 4.
• Root moduli. 2010 Exam 2 Section B Q5b demanded the answer in the form a^{1/n} cis(bπ/c) with integer a, b, c, n; [RPT10 E2]: "A few students did not use the required form for a^{1/n}, writing 2^{1/3}·5^{5/6} instead of 200^{1/6}."

The discipline: in Exam 1, never convert a surd to a decimal, never leave a surd in a denominator, and write the expansion line in full before cancelling — the marks in "show that" parts are for the expansion, not the conclusion.

---

6. A worked method sheet — the ten highest-yield types

Each method below is stated in its technology-free form, because that is the binding constraint: Examination 1 complex-number questions have a median full-mark rate of 48.5%, the lowest slice of the topic. Where CAS changes the approach on Examination 2, that is noted separately.

---

M1 — Cartesian → polar

Trigger phrases: "Express … in polar form"; "Show that … = r cis(θ)"; any question that goes on to ask for a power or a root.

1. r = √(x² + y²). Simplify the surd completely.
2. Reference angle φ = tan⁻¹(|y|/|x|).
3. Sketch the point. Assign θ by quadrant: Q1 = φ; Q2 = π − φ; Q3 = −(π − φ); Q4 = −φ.
4. Check −π < θ ≤ π.
5. Write r cis(θ).

Worked (from 2019 Exam 1 Q7a, 81%): 3 − √3 i. r = √(9 + 3) = √12 = 2√3. φ = tan⁻¹(√3/3) = π/6. Fourth quadrant, so θ = −π/6. Hence 3 − √3 i = 2√3 cis(−π/6). ∎

Exam 2 note: CAS will convert, but it may return an argument outside (−π, π] or in degrees. Always check the interval yourself.

---

M2 — De Moivre to a power, answer in a + bi

Trigger phrases: "Evaluate … giving your answer in the form a + bi"; "Write … in the form a + bi".

1. Convert every factor to polar (M1).
2. zⁿ = rⁿ cis(nθ); for a quotient, divide moduli and subtract arguments.
3. Reduce the resulting argument into (−π, π] by adding or subtracting multiples of 2π.
4. Convert back: r cis(θ) = r cos(θ) + i r sin(θ) using exact values.
5. Check the answer's quadrant against the sign of your θ.

Worked (from 2018 Exam 1 Q2b, 37%): (√3 − i)¹⁰/(1 + i)¹². √3 − i = 2cis(−π/6); 1 + i = √2 cis(π/4). Numerator = 2¹⁰ cis(−10π/6) = 1024 cis(−5π/3). Denominator = (√2)¹² cis(3π) = 64 cis(π). Quotient = 16 cis(−5π/3 − π) = 16 cis(−8π/3) = 16 cis(−8π/3 + 2π) = 16 cis(−2π/3). Hmm — check: −8π/3 + 2π = −2π/3 ✓, in range. So the answer is 16(cos(−2π/3) + i sin(−2π/3)) = 16(−½ − (√3/2)i) = −8 − 8√3 i. ∎ (This matches the [QJSON] answer field, −8 − 8√3 i.)

The two marks that go missing: step 3 (unreduced argument) and step 4 (leaving the answer in cis form when a + bi was demanded).

---

M3 — nth roots of a complex number

Trigger phrases: "Find the cube roots of …"; "Find all solutions of zⁿ = w"; "Find all distinct complex numbers z such that z³ = z₁".

1. Write w = r cis(θ) with θ principal (M1).
2. Modulus of each root: r^{1/n} — a real nth root, left in surd or index form.
3. First root: cis(θ/n). Remaining roots: step the argument by 2π/n.
4. Reduce every argument into (−π, π] if principal values are demanded.
5. If Cartesian form is demanded, convert each root with exact values.
6. Count your roots. There must be exactly n.

Worked (from 2015 Exam 1 Q4a, 40%): z³ = 8i. 8i = 8 cis(π/2). Modulus of each root = 2. Arguments: π/6, π/6 + 2π/3 = 5π/6, π/6 + 4π/3 = 3π/2 → 3π/2 − 2π = −π/2. So z = 2cis(π/6), 2cis(5π/6), 2cis(−π/2) = √3 + i, −√3 + i, −2i. ∎

Geometric check (free marks): the n roots are the vertices of a regular n-gon on the circle of radius r^{1/n}. If your roots are not equally spaced, you have made an error. If w is real and positive, z = r^{1/n} is always one of them.

---

M4 — Roots of unity

Trigger phrases: "List the other roots of zⁿ − 1 = 0"; "plot and label the points that represent all the roots".

1. The roots are cis(2kπ/n), k = 0, 1, …, n − 1. z = 1 is always a root — this is the one most often omitted.
2. In principal form, use k = −⌊n/2⌋ … ⌊n/2⌋.
3. They lie on the unit circle, equally spaced by 2π/n, symmetric about the real axis.
4. Their sum is 0 (the coefficient of z^{n−1} is zero). Their product is (−1)^{n+1}.
5. zⁿ − 1 = (z − 1)(z^{n−1} + z^{n−2} + … + z + 1).

Why (4) and (5) matter: they are the machinery behind 2023 Exam 2 Section B Q2fii (7%). Pairing conjugate roots in the sum gives 1 + 2Σ_{k=1}^{3} cos(2kπ/7) = 0, hence cos(2π/7) + cos(4π/7) + cos(6π/7) = −1/2.

---

M5 — Conjugate root theorem, quadratic factor, and the cofactor

Trigger phrases: "Given that z = α is a solution … find all other solutions"; "write down a quadratic factor"; "find the values of a and b".

1. Check the coefficients are real. If any is not, stop — use grouping or the factor theorem instead (M6).
2. The conjugate ᾱ is also a root.
3. Quadratic factor: z² − 2Re(α)z + |α|².
4. Cofactor: equate coefficients, not long division. Write P(z) = (z² − 2Re(α)z + |α|²)(z² + bz + c) (or (…)(dz + e) for a cubic), expand the leading and constant terms first, then one middle term.
5. Solve the cofactor by completing the square.
6. Answer with solutions if the question says "solve" and factors if it says "factorise". Never mix them.

Worked (from 2025 Exam 1 Q8b–c, 61% then 42%): f(z) = z⁴ − 6z² + 25, given z₁ = 1 + 2i. Conjugate 1 − 2i. Quadratic factor z² − 2z + 5. Cofactor: z⁴ − 6z² + 25 = (z² − 2z + 5)(z² + bz + c). Constant: 5c = 25 ⟹ c = 5. Coefficient of z³: b − 2 = 0 ⟹ b = 2. So the second factor is z² + 2z + 5, giving z = −1 ± 2i. ∎

[RPT25 E1] on this exact question: "Students who used comparison of coefficients to find the quadratic factor were generally more successful than those who used long or synthetic division."

---

M6 — Solving over C when the coefficients are not real

Trigger: any i in a coefficient.

1. Do not use the conjugate root theorem.
2. Try grouping first. 2007 Exam 1 Q2: z³ − (√5 − i)z² + 4z − 4√5 + 4i = z²(z − (√5 − i)) + 4(z − (√5 − i)) = (z − (√5 − i))(z² + 4). 2024 Exam 1 Q1: 3z³ + 2iz² + 3z + 2i = z²(3z + 2i) + (3z + 2i) = (3z + 2i)(z² + 1).
3. Otherwise use the factor theorem with the given root and divide.
4. For a quadratic with a non-real coefficient, complete the square: z² + 2biz + c = (z + bi)² + c + b². 2022 Exam 1 Q1: z² + 6iz − 25 = (z + 3i)² − 16, so z = −3i ± 4.
5. If the discriminant is non-real, take its square root via polar form (M7).

---

M7 — Square root of a complex number (two routes)

Trigger: z² = w with w non-real; a quadratic over C whose discriminant is non-real; z⁴ + bz² + c = 0.

Route A (polar — preferred by the reports). Write w = r cis(θ); then z = √r cis(θ/2) and √r cis(θ/2 − π). Convert with exact values.

Route B (Cartesian). Set (x + iy)² = p + qi, giving x² − y² = p and 2xy = q. Solve the pair. Slower, and [RPT13 E1] records that "most who used this approach struggled with the algebra."

Worked (from 2013 Exam 1 Q8, 23%): z⁴ − 2z² + 4 = 0 ⟹ z² = (2 ± √(4 − 16))/2 = 1 ± √3 i. Now 1 + √3 i = 2 cis(π/3), so z = √2 cis(π/6) and √2 cis(π/6 − π) = √2 cis(−5π/6). Similarly 1 − √3 i = 2cis(−π/3) gives z = √2 cis(−π/6) and √2 cis(5π/6). Converting: z = ±(√6/2 + (√2/2)i) and ±(√6/2 − (√2/2)i) — four solutions, as required. ∎

Checkpoint: a quartic has four roots. [RPT13 E1] records z = ±√(1 ± √3 i) given as a final answer — that is not Cartesian form and scores nothing for the last two marks.

---

M8 — |z − a| = |z − b| to a Cartesian line

Trigger phrases: "Show that the cartesian equation of L is …"; "Express the relation |z − u| = |z − v| in the form y = mx + c".

Route A (algebraic — the majority route, and the safest for a "show that"). 1. Put z = x + iy. 2. |(x − a₁) + (y − a₂)i| = |(x − b₁) + (y − b₂)i|. 3. Square both sides. Write the full expansion. 4. x² and y² cancel. Collect into y = mx + c.

Route B (geometric — faster, accepted, useful as a check). 1. Midpoint of a and b. 2. Gradient of ab is m₁; the line has gradient −1/m₁. 3. Write the line through the midpoint.

Worked (from 2025 Exam 2 Section B Q2bi, 74.5%): show that |z − 2i| = |z − √3 − i| gives y = √3 x.

Algebraic. x² + (y − 2)² = (x − √3)² + (y − 1)² ⟹ x² + y² − 4y + 4 = x² − 2√3 x + 3 + y² − 2y + 1 ⟹ −4y + 4 = −2√3 x − 2y + 4 ⟹ −2y = −2√3 x ⟹ y = √3 x. ∎

Geometric (the route [RPT25 E2] sets out). The two points are (0, 2) and (√3, 1). Midpoint (√3/2, 3/2); gradient of the segment = (1 − 2)/(√3 − 0) = −1/√3; perpendicular gradient = √3; the line through the midpoint with gradient √3 is y − 3/2 = √3(x − √3/2), i.e. y = √3 x. ∎

[RPT25 E2]: "This was a 'show that' question which requires a full algebraic or a geometric approach… The algebraic approach is to equate the magnitudes of the complex expressions and then expand the brackets and simplify. The geometric approach required finding the midpoint and the gradient of the line segment."

Checkpoint: the line must pass through the midpoint of a and b. Substitute it. This catches the "most common error … a negative gradient" that [RPT16 E2] names.

---

M9 — The three circle forms, and the ray

Ray procedure: 1. z₀ is the point of emanation; mark it with an open circle. 2. Direction: θ measured anticlockwise from the positive real direction at z₀. 3. Cartesian rule: y − y₀ = tan(θ)(x − x₀), plus a domain: x > x₀ if −π/2 < θ < π/2; x < x₀ if |θ| > π/2; and the degenerate cases x = x₀, y > y₀ for θ = π/2 and x = x₀, y < y₀ for θ = −π/2. 4. To find θ from two points: θ = Arg(z₁ − z₀), principal value.

Worked (from 2024 Exam 2 Section B Q2dii, 56%): the ray originates at z₀ = 2 − i and passes through −2 + 3i. z₁ − z₀ = −4 + 4i = 4√2 cis(3π/4). So the equation is Arg(z − (2 − i)) = 3π/4. ∎ [RPT24 E2]: "The most common error was to quote the argument as −π/4."

---

M10 — Areas in the complex plane

Trigger phrases: "Find the area of the minor/major segment"; "Find the area of the shaded region"; "Find the area in the first quadrant enclosed by …".

Four formulas, all from [FS]'s Mensuration block or elementary geometry:

Procedure: 1. Identify the two intersection points on the circle. 2. Find the central angle they subtend — the angle at the centre, which is not the ray's argument and not the inscribed angle. Use the cosine rule on the triangle (centre, P, Q), or the dot product, or symmetry. 3. Decide minor or major by reading the question's adjective. 4. Leave the answer exact unless told otherwise.

Worked (from 2024 Exam 2 Section B Q2e, 52%): circle |z − (1 + 2i)| = 2, so r = 2. [RPT24 E2]: "As the circle has radius 2 and the minor segment has angle π/2, its area is ½ · 4 · (π/2 − sin(π/2)) = 2(π/2 − 1) = π − 2." ∎

Do not integrate. [RPT16 E2]: "a larger number set up elaborate definite integrals to find the area, occasionally successfully, but this was not an efficient approach." [RPT09 E2]: "Often elaborate approaches were set up to solve this simple problem."

---

Appendix — the quick-reference error checklist

Run this over every complex-number answer before moving on. Each line corresponds to at least one separator in §4.

Form - [ ] Is the answer in the form the question named — a + ib, r cis(θ), x + iy, "as a product of linear factors"? - [ ] Is it exact? No decimals unless "correct to … decimal places" appears. - [ ] If the question said "solve", are these solutions; if "factorise", are these factors?

Argument - [ ] Is every argument in (−π, π]? - [ ] Does the argument's quadrant match the signs of the real and imaginary parts? - [ ] If the answer is a set of n values, is the parameter defined (k ∈ Z)?

Completeness - [ ] For a degree-n equation, are there n roots (counted with multiplicity)? - [ ] For nth roots, are all n given and equally spaced? - [ ] For an intersection, have you found all intersection points (a circle and a line meet twice; |z + z̄| = |z − z̄| is two lines)? - [ ] For ± from a square root, are both branches kept?

Conjugates - [ ] Are the polynomial's coefficients real? If not, no conjugate pairs are forced. - [ ] Is the conjugate ᾱ, not −α?

Diagrams - [ ] Points marked as dots or crosses (not line segments from the origin) and labelled? - [ ] Points on the correct circle — check the modulus against the grid? - [ ] Ray endpoints open; lines drawn full-length; rays not extended backwards? - [ ] Axis intercepts labelled when asked? - [ ] Shading confined to the intersection, corner points included? - [ ] Circle drawn smoothly — [RPT24 E2]: "should not have a pointed shape"?

"Show that" - [ ] Does the working move forward to the given result, not backwards from it? - [ ] Is every algebraic step visible, especially the full expansion? - [ ] Have you avoided substituting the given answer to "verify" it, unless the word is "verify"?