2.2 Bar Charts, Grouped Bars & Component/Stacked Graphs

2.2 Bar Charts, Grouped Bars & Component/Stacked Graphs

[!NOTE] Psychometric Focus: Bar charts evaluate your visual-to-numerical translation speed. On the CSNT, you will encounter multiple bar variations—simple, grouped, stacked, and 100% component charts. The core cognitive challenge is rapidly calibrating axis gridlines, distinguishing relative proportions from absolute volumes, and resisting visual illusions caused by truncated scales.

Bar charts represent categorical or discrete data using rectangular bars whose lengths or heights are proportional to the values they represent. Unlike continuous line graphs, bar charts separate categories with discrete whitespace. In UK government reporting, bar charts illustrate operational performance, workforce diversity statistics, regional budget allocations, and target-versus-actual analyses.

---

Typology of Bar Charts in Civil Service Testing

1. Simple Horizontal vs. Vertical Bar Charts

• Vertical Bar Charts (Column Charts): Best suited for discrete chronological intervals (e.g., quarterly expenditures) or small sets of categorical entities (e.g., five departments).
• Horizontal Bar Charts: Preferred when category names are lengthy (e.g., "Crown Prosecution Service", "Department for Environment, Food & Rural Affairs") or when items are arranged in rank order (e.g., highest to lowest caseload).

2. Grouped (Clustered) Bar Charts

Grouped bar charts display two or more sub-bars within each categorical grouping, sharing a common baseline axis. For example, within each department, you might see side-by-side bars for 2024 Spend, 2025 Spend, and 2026 Budget Target.

• Key Technique: Always reference the visual legend before reading values. Verify whether clusters are grouped by Department (with years side-by-side) or grouped by Year (with departments side-by-side).

3. Stacked (Component) Bar Charts

In a stacked bar chart, each bar represents a cumulative total, subdivided into shaded or colored segments that reflect component categories. For instance, a total recruitment bar may be split into Direct Entry and Fast Stream recruits.

• The "Floating Segment" Principle: While the bottom segment begins at zero and can be read directly from the vertical axis, every subsequent segment "floats" above earlier segments. To determine the absolute value of a floating segment, you must subtract its lower boundary from its upper boundary:

$\text{Segment Value} = \text{Upper Axis Value} - \text{Lower Axis Value}$

Reading the upper boundary directly from the axis is among the most frequent calculation errors made by test-takers on stacked bar questions.

4. 100% Component (Normalized) Bar Charts

A 100% component bar chart normalizes all bars to an identical $100%$ height. The segments illustrate the percentage share of each component within its respective category, completely independent of absolute scale.

• Core Rule: You cannot determine absolute quantities from a 100% component chart alone unless the absolute total volume for each bar is explicitly provided in an accompanying data label or footnote. A department with a $40%$ segment in a 100% chart may have far fewer actual caseworkers than a department with a $20%$ segment if the latter department is significantly larger in total size.

---

Departmental Scenario: Recruitment Targets vs. Actual Intakes

Consider the following data reflecting the annual intake of new personnel across five major UK government departments, illustrating how grouped and stacked visual data is structured in psychometric items.

Table: Annual Intake vs. Target by Department (2025/26)

Deconstructing the Data:

• Direct Entry vs. Fast Stream Proportions: Across the overall civil service cohort, Direct Entry accounted for $2,690 / 3,985 = 67.50%$, while the Civil Service Fast Stream accounted for $1,295 / 3,985 = 32.50%$.
• Departmental Fast Stream Ratios: In HMRC, the Fast Stream intake was exactly $315 / 945 = 33.33%$ ($1/3$ of intake). In Defra, Fast Stream intake was $140 / 420 = 33.33%$. In DWP, Fast Stream intake was $410 / 1,260 = 32.54%$.
• Target Achievement Variance: DWP and HMRC both exceeded their recruitment targets by exactly $+5.0%$, whereas the Ministry of Justice (MoJ) suffered the steepest percentage deficit ($-7.69%$), falling $50$ recruits short of its $650$ target.

---

Visual Traps and Psychometric Pitfalls

1. Truncated Axes (Non-Zero Origins)

A truncated axis begins at a value greater than zero (e.g., starting at $400$ instead of $0$). This compresses the visual baseline and drastically exaggerates relative differences. For example, if Department A achieves $420$ recruits and Department B achieves $460$ recruits:

• On a true baseline ($0$ to $500$), Bar B is only $460 / 420 = 1.095$ times ($9.5%$) taller than Bar A.
• On a truncated axis starting at $400$, Bar A has a visible height of $420 - 400 = 20$ units, while Bar B has a visible height of $460 - 400 = 60$ units. Visually, Bar B appears three times ($300%$) as tall as Bar A!

[!WARNING] Never Judge by Eye on Truncated Charts: Always read the numeric values from the axis tick marks or data labels. Never estimate proportional magnitude based on physical bar height when the origin does not begin at zero.

2. Dual-Axis Misalignment

Some complex charts present two distinct vertical axes: a left y-axis displaying absolute volumes (e.g., Headcount) and a right y-axis displaying financial amounts (e.g., Budget in £m) or percentages. Always trace each data series to its corresponding axis before taking readings.

Worked Step-by-Step Examples

Worked Example 1: Segment Extraction from a Stacked Bar

Scenario: In the recruitment chart above, what is the ratio of Direct Entry recruits to Fast Stream recruits across the Department for Work and Pensions (DWP), expressed in lowest integer terms?

Step 1: Extract the segment values for DWP

• Direct Entry segment = $850$
• Fast Stream segment = $410$

Step 2: Formulate the unsimplified ratio $\text{Ratio} = 850 : 410$

Step 3: Simplify by dividing by the greatest common divisor

• Divide both sides by 10: $85 : 41$.
• 41 is a prime number and does not divide 85 ($85 / 41 = 2.073$). Therefore, the ratio in lowest whole numbers is $85 : 41$.

---

Worked Example 2: Comparing Proportions Across Unequal Bases

Scenario: What is the difference in percentage points between the Fast Stream share of total intake in the Home Office versus the Ministry of Justice (MoJ)?

Step 1: Calculate the Fast Stream percentage in the Home Office $\text{Home Office Total Actual Intake} = 520 + 240 = 760$ $\text{Fast Stream \%} = \frac{240}{760} \times 100 = 31.579\% \approx 31.58\%$

Step 2: Calculate the Fast Stream percentage in the MoJ $\text{MoJ Total Actual Intake} = 410 + 190 = 600$ $\text{Fast Stream \%} = \frac{190}{600} \times 100 = 31.667\% \approx 31.67\%$

Step 3: Compute the percentage point difference $\text{Difference} = 31.667\% - 31.579\% = 0.088\text{ percentage points}$

Despite the Home Office having a much higher absolute number of Fast Stream recruits ($240$ vs $190$), its proportionate share within its own departmental intake is virtually identical to that of the MoJ (a negligible difference of less than $0.1$ percentage points).