1.3 Numerical Reasoning: Ratios, Percentages & Data Interpretation

Numerical Reasoning: Ratios, Percentages & Data Interpretation

Quick Summary: The EPSO numerical reasoning examination evaluates a candidate's speed, precision, and analytical judgment in interpreting financial, economic, and operational data. Operating under a pacing budget of approximately two minutes per question, candidates must master mental estimation, percentage manipulations, index numbers, and weighted averages while utilizing a standard on-screen four-function calculator.

---

Test Format, Timing & Interface Constraints

The numerical reasoning test presents quantitative scenarios mirroring business, statistical, and public administration contexts. Key test parameters include:

• Item Count: 10 questions.
• Allocated Time: 20 minutes.
• Pacing Budget: Exactly 120 seconds (2 minutes) per question.
• Scoring Architecture: Combined with Abstract Reasoning for an aggregate threshold of 10 out of 20 points. There is no separate pass mark for numerical reasoning alone.
• On-Screen Interface: Follow the invitation and current EPSO instructions. The 2026 platform guidance describes an on-screen calculator and, when enabled, a digital scratchpad; personal paper, whiteboards and external calculators should not be assumed permitted.

---

Core Mathematical Competencies for Audit Candidates

While the mathematical concepts tested are grounded in secondary-school arithmetic, the questions require multi-step data extraction from complex tables, double-axis charts, and clustered graphs. Candidates must master six fundamental competencies:

1. Percentage Changes and Growth Multipliers

The fundamental formula for relative percentage change between an initial value ($V_0$) and a final value ($V_1$) is:

$\Delta \% = \frac{V_1 - V_0}{V_0} \times 100\% = \left( \frac{V_1}{V_0} - 1 \right) \times 100\%$

To maximize speed, candidates should convert percentage changes into decimal multipliers:

• An increase of $8.5\%$ corresponds to a multiplier of $1 + 0.085 = 1.085$.
• A decrease of $14\%$ corresponds to a multiplier of $1 - 0.14 = 0.86$.

The Compounding and Asymmetry Trap: Percentage changes are non-commutative and asymmetric. An increase of $25\%$ followed by a decrease of $20\%$ returns the value to its baseline ($1.25 \times 0.80 = 1.00$). However, an increase of $20\%$ followed by a decrease of $20\%$ results in a net loss of $4\%$ ($1.20 \times 0.80 = 0.96$).

2. Percentage Points versus Relative Percentage Change

A frequent distractor trap exploits the confusion between absolute changes in rates (percentage points) and relative changes in rates (percentage change):

• If an audit error rate increases from $2.0\%$ in Year 1 to $3.0\%$ in Year 2:
  • The absolute increase is $3.0\% - 2.0\% = 1.0\text{ percentage point}$ (pp).
  • The relative increase is $\frac{3.0 - 2.0}{2.0} \times 100\% = 50.0\%$.

3. Proportions, Segment Shares, and Ratios

Calculating the share of a specific expenditure line relative to an aggregate total requires isolating the component numerator and dividing by the composite denominator:

$\text{Share} = \frac{\text{Sub-component}}{\sum \text{All Components}} \times 100\%$

When calculating part-to-part ratios (such as administrative expenditure to operational expenditure), ensure that the terms are not inverted.

4. Index Numbers and Deflators

Index numbers benchmark statistical data relative to a base period assigned a value of $100$:

$\text{Index}_t = \frac{\text{Value}_t}{\text{Value}_{\text{base}}} \times 100$

Calculating Changes Between Non-Base Periods: To find the percentage change between two non-base index years (e.g., Year 3 at index $115$ and Year 5 at index $138$), do not simply subtract the index points ($138 - 115 = 23\%$ is incorrect). You must calculate the relative change between the two indices:

$\Delta \% = \frac{138 - 115}{115} \times 100\% = \frac{23}{115} \times 100\% = 20.0\%$

5. Weighted Averages in Audit Testing

When combining figures across populations of differing sizes (such as calculating the aggregate error rate across three operational programs with different total expenditures), a simple arithmetic mean yields an erroneous result. You must apply a weighted average:

$\bar{X}_w = \frac{\sum_{i=1}^n (w_i \cdot x_i)}{\sum_{i=1}^n w_i}$

where $w_i$ represents the population weight (e.g., total program spending) and $x_i$ represents the observed metric (e.g., detected error percentage).

6. Budget Variance & Absorption Rates

Public sector audit questions frequently assess budgetary execution dynamics:

• Budget Variance: $\text{Variance} = \text{Actual Expenditure} - \text{Committed / Budgeted Amount}$.
• Budget Absorption Rate: $\text{Absorption} = \frac{\text{Payments Executed}}{\text{Commitment Appropriations}} \times 100\%$.
• Decommitment: Unspent funds subject to cancellation under EU $N+2$ or $N+3$ decommitment rules.

---

High-Speed Calculation Techniques and Heuristics

With only 120 seconds per question, entering 10-digit numbers into an on-screen calculator invites transcription errors and consumes valuable time. Implement these heuristics:

Loading diagram...

1. Option Spread Inspection: Before touching the calculator, look at the four answer options. If the options are widely dispersed (e.g., $12.4\%$, $21.8\%$, $35.2\%$, $49.0\%$), rough rounding to two significant digits will reveal the correct answer in 20 seconds.
2. Multiplier Chaining: For compounding adjustments, multiply the decimal multipliers in sequence rather than computing intermediate sub-totals. For example, to find a value that grew by $5\%$ in Year 1 and dropped by $8\%$ in Year 2: $\text{Final} = \text{Base} \times 1.05 \times 0.92 = \text{Base} \times 0.966$.
3. Boundary Value Checks: A weighted average of two values (e.g., $4.2\%$ and $8.6\%$) must strictly lie between $4.2\%$ and $8.6\%$. Any option outside this interval can be instantly eliminated.
4. Cross-Multiplication for Fraction Comparison: When comparing fractions to determine which is larger (e.g., $\frac{325}{1,480}$ vs $\frac{410}{1,850}$), cross-multiply: $325 \times 1,850 = 601,250$ versus $1,480 \times 410 = 606,800$. Because $606,800 > 601,250$, the second fraction is larger.

---

Essential Quantitative Formulas for Audit Candidates

---

Comprehensive Worked Numerical Case

Public Finance Scenario: Cohesion Policy Fund Execution

The following table illustrates multi-year operational data for four Member States under the European Regional Development Fund (ERDF) across two consecutive programming years:

Practical Problem 1: Comparing Budget Absorption Rates

Question: Which Member State demonstrated the highest percentage point increase in its budget absorption rate (Executed Payments / Initial Commitment) between Year 1 and Year 2?

Step-by-Step Calculation:

1. Calculate Year 1 Absorption Rate ($\frac{\text{Executed}}{\text{Commitment}}$):
   • State Alpha: $\frac{320}{400} = 80.0\%$
   • State Beta: $\frac{175}{250} = 70.0\%$
   • State Gamma: $\frac{540}{600} = 90.0\%$
   • State Delta: $\frac{90}{150} = 60.0\%$
2. Calculate Year 2 Absorption Rate ($\frac{\text{Executed}}{\text{Commitment}}$):
   • State Alpha: $\frac{405}{450} = 90.0\%$
   • State Beta: $\frac{240}{300} = 80.0\%$
   • State Gamma: $\frac{520}{650} = 80.0\%$
   • State Delta: $\frac{150}{200} = 75.0\%$
3. Calculate Percentage Point (pp) Difference ($\text{Year 2} - \text{Year 1}$):
   • State Alpha: $90.0\% - 80.0\% = +10.0\text{ pp}$
   • State Beta: $80.0\% - 70.0\% = +10.0\text{ pp}$
   • State Gamma: $80.0\% - 90.0\% = -10.0\text{ pp}$
   • State Delta: $75.0\% - 60.0\% = \mathbf{+15.0\text{ pp}}$

Conclusion: State Delta experienced the greatest increase ($+15.0$ percentage points).

Practical Problem 2: Multi-Sample Weighted Error Rate

Question: Across State Alpha and State Gamma combined in Year 2, what was the composite observed irregularity rate in the audited sample?

Step-by-Step Calculation:

1. Determine combined audited sample expenditure (denominator):
   • Alpha Sample $+$ Gamma Sample $= 12,000\text{ EURk} + 20,000\text{ EURk} = 32,000\text{ EURk}$ (EUR 32,000,000).
2. Determine combined detected irregularities (numerator):
   • Alpha Irregularities $+$ Gamma Irregularities $= 480\text{ EURk} + 1,200\text{ EURk} = 1,680\text{ EURk}$ (EUR 1,680,000).
3. Calculate composite error rate:
   • $\text{Composite Rate} = \frac{1,680}{32,000} = \frac{168}{3,200} = \frac{21}{400} = \mathbf{5.25\%}$.

Verification via individual rates: Alpha rate is $\frac{480}{12,000} = 4.0\%$. Gamma rate is $\frac{1,200}{20,000} = 6.0\%$. Because Gamma accounts for $\frac{20}{32} = 62.5\%$ of the combined sample, the weighted average must be closer to $6.0\%$ than $4.0\%$. Indeed, $0.375 \times 4.0\% + 0.625 \times 6.0\% = 1.5\% + 3.75\% = 5.25\%$. Matches perfectly.