7.1 Number Sequences & Pattern Recognition

Section 7.1: Number Sequences & Pattern Recognition

The Role of Sequence Recognition in Civil Service Testing

In public administration, civil servants frequently track dynamic data streams over chronological intervals: monthly benefit claims submitted to the Department for Work and Pensions (DWP), phased quarterly milestone outlays on major infrastructure projects, or cumulative recruitment intakes across multi-stage civil service campaigns. Aptitude testing bodies (including modern adaptive CSNT assessments) employ numerical sequence questions to measure an applicant's inductive reasoning, pattern recognition agility, and ability to detect structural mathematical rules.

On the Civil Service Numerical Test, sequence questions present an ordered string of numbers with one or two missing values (often denoted by a question mark or blank bracket). Your objective is to uncover the governing rule that transforms each term into the next, and then extrapolate that rule to determine the missing value. While elementary tests often feature simple addition or multiplication steps, the CSNT regularly incorporates higher-order patterns: second-tier differences, interleaved alternating sub-series, multi-operation recursive loops, and Fibonacci-style additive structures.

---

Typology of Numerical Sequences

1. Arithmetic Progressions (First-Order Constant Differences)

An arithmetic progression (AP) is the most fundamental sequence type. The numerical difference between any two consecutive terms is a fixed constant, known as the common difference ($d$):

$d = x_n - x_{n-1}$

The general $n$-th term is formulated as:

$x_n = a + (n - 1)d$

Where $a$ is the initial term and $d$ is the constant difference.

• If $d > 0$, the sequence increases monotonically (e.g., $14, 19, 24, 29, 34, \dots$ where $d = +5$).
• If $d < 0$, the sequence decreases monotonically (e.g., $82, 75, 68, 61, 54, \dots$ where $d = -7$).

Civil Service Application: Routine intake allocations where an executive agency onboards a fixed tranche of apprentices each month.

2. Geometric Progressions (Constant Common Ratios)

In a geometric progression (GP), each successive term is generated by multiplying the preceding term by a constant factor, termed the common ratio ($r$):

$r = \frac{x_n}{x_{n-1}}$

The general $n$-th term is formulated as:

$x_n = a \cdot r^{n-1}$

• If $r > 1$, the values grow exponentially (e.g., $6, 18, 54, 162, 486, \dots$ where $r = 3$).
• If $0 < r < 1$, the values contract exponentially toward zero (e.g., $640, 320, 160, 80, 40, \dots$ where $r = 0.5$).
• If $r < 0$, the signs of the terms alternate between positive and negative values (e.g., $5, -15, 45, -135, \dots$ where $r = -3$).

Civil Service Application: Modelling viral public health caseload trajectories, exponential compounding in long-term debt servicing, or rapid technological adoption rates across citizen portals.

3. Second-Order (Two-Tier) Difference Sequences

When the first differences between consecutive terms are not constant, but instead increase or decrease by a steady increment, the sequence is a second-order difference sequence. This behavior reveals an underlying quadratic equation of the form:

$x_n = a n^2 + b n + c$

In any quadratic sequence, the second difference ($D_2$) is mathematically constant and equals $2a$:

$D_2 = \Delta_{1, k+1} - \Delta_{1, k} = 2a = \text{constant}$

Illustrative Structural Breakdown:

Consider the sequence: $4, 9, 18, 31, 48, 69, [?]$

Let us compute the first differences (Tier 1): $\Delta_{1, 1} = 9 - 4 = +5$ $\Delta_{1, 2} = 18 - 9 = +9$ $\Delta_{1, 3} = 31 - 18 = +13$ $\Delta_{1, 4} = 48 - 31 = +17$ $\Delta_{1, 5} = 69 - 48 = +21$

Now compute the second differences (Tier 2): $D_2 = 9 - 5 = +4$ $D_2 = 13 - 9 = +4$ $D_2 = 17 - 13 = +4$ $D_2 = 21 - 17 = +4$

The second difference is constant at $+4$. Therefore, the next first difference must be $21 + 4 = +25$. Adding this difference to the latest term yields the missing term: $69 + 25 = 94$.

4. Alternating and Interleaved Sequences

An interleaved sequence (also known as a dual or composite sequence) blends two independent sub-sequences into a single series by alternating between positions:

• Odd positions ($n = 1, 3, 5, 7, \dots$) follow Rule A.
• Even positions ($n = 2, 4, 6, 8, \dots$) follow Rule B.

Key Visual Diagnostic: The numbers do not move in a consistent direction; instead, they oscillate up and down irregularly, or feature sudden jumps that violate local progression logic.

Illustrative Example:

Consider the sequence: $12, 40, 15, 36, 18, 32, 21, 28, [?]$

Separating the terms by index position reveals two completely independent progressions:

• Odd-indexed terms ($1^{\text{st}}, 3^{\text{rd}}, 5^{\text{th}}, 7^{\text{th}}, 9^{\text{th}}$): $12 \xrightarrow{+3} 15 \xrightarrow{+3} 18 \xrightarrow{+3} 21 \xrightarrow{+3} [?]$. This is an arithmetic progression with $d = +3$. The $9^{\text{th}}$ term is $21 + 3 = 24$.
• Even-indexed terms ($2^{\text{nd}}, 4^{\text{th}}, 6^{\text{th}}, 8^{\text{th}}$): $40 \xrightarrow{-4} 36 \xrightarrow{-4} 32 \xrightarrow{-4} 28$. This is an arithmetic progression with $d = -4$.

Because the missing value occupies the $9^{\text{th}}$ (odd) position, the answer is 24.

5. Multi-Operation Recursive Patterns

In a multi-operation sequence, a single series applies a cyclic combination of distinct arithmetic operations between adjacent terms. Rather than a uniform $+d$ or $\times r$, the transformation alternates across a repeating operational cycle, such as:

• $(\times 2, +3, \times 2, +3, \dots)$
• $(+4, \times 3, +4, \times 3, \dots)$
• $(\times 2, -1, \times 3, -2, \times 4, -3, \dots)$

Illustrative Example:

Consider the sequence: $3, 6, 10, 20, 24, 48, 52, [?]$

Let us analyze the transitions:

• $3 \rightarrow 6$: $\times 2$
• $6 \rightarrow 10$: $+ 4$
• $10 \rightarrow 20$: $\times 2$
• $20 \rightarrow 24$: $+ 4$
• $24 \rightarrow 48$: $\times 2$
• $48 \rightarrow 52$: $+ 4$

The pattern is a repeating two-step operational cycle: $[\times 2, +4]$. The next required transformation is $\times 2$. Thus, the missing term is $52 \times 2 = 104$.

6. Additive and Fibonacci-Type Sequences

In a Fibonacci-style sequence, each term is generated not by operating on the immediate predecessor with an external constant, but by summing the preceding two (or more) terms:

$x_n = x_{n-1} + x_{n-2}$

Standard Fibonacci series: $1, 1, 2, 3, 5, 8, 13, 21, 34, \dots$

On the CSNT, exam developers frequently introduce modified variations:

• Arbitrary starting seeds: e.g., starting with $3$ and $7$: $3, 7, 10, 17, 27, 44, 71, \dots$
• Additive plus constant: $x_n = x_{n-1} + x_{n-2} + k$
• Subtractive variations: $x_n = x_{n-1} - x_{n-2}$

---

The 5-Step Systematic Diagnostic Workflow

The CSNT is untimed, but unsystematic trial-and-error on a sequence item still drains focus and tempts a guess — and on an adaptive test a guessed wrong answer costs you difficulty. Adhere strictly to this ordered diagnostic workflow:

[Given Sequence: x_1, x_2, x_3, x_4, x_5, ...]
         │
         ▼
[Step 1: Calculate First Differences (Δ_1 = x_{k+1} - x_k)]
  ├── Constant? ───────────► Arithmetic Progression (Done!)
  └── Not constant? ───────► Proceed to Step 2
         │
         ▼
[Step 2: Calculate Consecutive Ratios (r = x_{k+1} / x_k)]
  ├── Constant? ───────────► Geometric Progression (Done!)
  └── Not constant? ───────► Proceed to Step 3
         │
         ▼
[Step 3: Calculate Second Differences (D_2 = Δ_{1, k+1} - Δ_{1, k})]
  ├── Constant? ───────────► Quadratic / 2nd-Order Pattern (Done!)
  └── Not constant? ───────► Proceed to Step 4
         │
         ▼
[Step 4: Check for Interleaved Series (Split Odd & Even Indices)]
  ├── Sub-patterns emerge? ─► Dual Interleaved Progression (Done!)
  └── No dual pattern? ────► Proceed to Step 5
         │
         ▼
[Step 5: Test Multi-Operation Loops or Additive Fibonacci Structures]
  ├── Repeating op pair? ──► Multi-Operation Recursive (Done!)
  └── Sum of prior terms? ─► Fibonacci-Type Progression (Done!)

[!TIP] Diagnostic Rapid Check: If terms grow very rapidly (e.g., from double digits to four digits in four steps), prioritize Step 2 (ratios) or Step 5 (multiplication loops). If terms fluctuate in direction (up, down, up, down), jump immediately to Step 4 (interleaved split).

---

Worked Civil Service Scenarios

Scenario 1: Department for Transport Rail Modernisation Capital Milestones

The Department for Transport (DfT) tracks quarterly capital expenditures (£ millions) allocated to an ongoing rail corridor electrification programme. The audited milestone disbursements for the first five quarters are recorded as follows:

$\text{Q1: } £15\text{m}, \quad \text{Q2: } £23\text{m}, \quad \text{Q3: } £35\text{m}, \quad \text{Q4: } £51\text{m}, \quad \text{Q5: } £71\text{m}$

If the underlying contractual milestone formula continues, what is the projected capital requirement for Quarter 6 (Q6) and Quarter 7 (Q7)?

Step-by-Step Solution via the Diagnostic Workflow:

1. Step 1 (First Differences): $\Delta_{1, 1} = 23 - 15 = +8$ $\Delta_{1, 2} = 35 - 23 = +12$ $\Delta_{1, 3} = 51 - 35 = +16$ $\Delta_{1, 4} = 71 - 51 = +20$ The first differences ($8, 12, 16, 20$) are not constant, ruling out a simple arithmetic progression.
2. Step 2 (Ratios): $23 / 15 \approx 1.533, \quad 35 / 23 \approx 1.522, \quad 51 / 35 \approx 1.457$ The ratio is not constant, ruling out a geometric progression.
3. Step 3 (Second Differences): $D_2 = 12 - 8 = +4$ $D_2 = 16 - 12 = +4$ $D_2 = 20 - 16 = +4$ The second difference is strictly constant at $+4$. This confirms a second-order quadratic spend progression.
4. Extrapolating Future Terms:
   • For Q6: The next first difference must be $20 + 4 = +24$. $\text{Q6 Outlay} = 71 + 24 = £95\text{m}$
   • For Q7: The subsequent first difference must be $24 + 4 = +28$. $\text{Q7 Outlay} = 95 + 28 = £123\text{m}$

The projected capital requirements are £95m for Q6 and £123m for Q7.

---

Scenario 2: Civil Service Fast Stream Headcount Intake Projections

Cabinet Office workforce planners model intake volumes across successive recruitment cohorts for a specialized digital transformation stream. The cohort intake numbers over the first six intakes are logged as:

$8, 16, 13, 26, 23, 46, [?], [?]$

What are the projected headcount intake numbers for the $7^{\text{th}}$ and $8^{\text{th}}$ recruitment cohorts?

Step-by-Step Solution via the Diagnostic Workflow:

1. Examine Progression Dynamics: The numbers rise, drop slightly, rise sharply, drop slightly, and rise sharply. This oscillating behavior immediately rules out pure AP or GP.
2. Test Interleaved Split (Step 4):
   • Odd terms: $8, 13, 23, \dots$ (differences: $+5, +10$)
   • Even terms: $16, 26, 46, \dots$ (differences: $+10, +20$) While a pattern is visible, let us check Step 5 for a cleaner direct operational rule.
3. Test Multi-Operation Loop (Step 5):
   • Term 1 to Term 2: $8 \times 2 = 16$
   • Term 2 to Term 3: $16 - 3 = 13$
   • Term 3 to Term 4: $13 \times 2 = 26$
   • Term 4 to Term 5: $26 - 3 = 23$
   • Term 5 to Term 6: $23 \times 2 = 46$ The operational rule is a consistent alternating cycle: $[\times 2, -3]$.
4. Extrapolating Terms 7 and 8:
   • The next operation after $\times 2$ is $-3$: $\text{Term 7} = 46 - 3 = 43$
   • The operation following $-3$ is $\times 2$: $\text{Term 8} = 43 \times 2 = 86$

The projected cohort intakes are 43 and 86 recruits respectively.

---

Sequence Summary Matrix