Data visualization should make a defined comparison easier to inspect. A chart can display its numbers accurately and still invite the wrong conclusion if the measure, denominator or scope is unclear. Begin with the question, then build a display that preserves the evidence needed to answer it.
This guide creates two original text bar charts from invented task records. It includes the complete input table, count and rate calculations, accessible text descriptions, a misleading-scale repair and two changed-input checks. No company dataset, dashboard deployment or business improvement is represented.
Define the question before choosing a chart
At a single fictional snapshot, the North room has fifteen completed tasks among twenty-five assigned tasks. The South room has eighteen completed tasks among forty assigned tasks. The room names are invented categories, not geographical populations or employee-performance rankings.
The first question is “Which room has more completed tasks in the supplied records?” The second is “Which has a larger completed share of its assigned tasks?” Those questions use different measures, so the guide gives each its own chart.
| Room | Supplied complete / assigned counts |
|---|---|
| North | 15 / 25 |
| South | 18 / 40 |
The input contains no task difficulty, staffing, duration or representative sampling information. A visual difference cannot fill those gaps. It cannot show why one count or share differs from another.
Keep preparing
Continue with Sarkari Resume Templates₹299 — coaching के एक महीने से काफ़ी सस्ता / far cheaper than a month of coachingCalculate the measures independently
South has three more completed tasks than North: 18 − 15 = 3. For completion share, North is 15 ÷ 25 = 60%, while South is 18 ÷ 40 = 45%. North's share is fifteen percentage points higher.
The count favours South; the share favours North. This is not a contradiction. South's supplied assigned total is larger. Count measures completed volume, while the defined rate describes the fraction completed out of each supplied assigned total.
Do not label either result “the better team.” The exercise does not establish comparable tasks, a performance target or a causal explanation. The supported conclusion is narrower: these particular counts and shares differ as shown.
Use narrow chart guidance
The UK's Office for National Statistics axes guidance recommends zero baselines for bar charts and consistent scales for comparable charts. Its guidance distinguishes bars from line and scatter displays, where a cropped axis can sometimes be appropriate. This guide uses original bar displays; it does not copy an ONS example or claim that every chart must start at zero.
ONS's chart-text guidance supports clear measures, units, scope and source information, with legible text. The displays below use explicit labels and accompanying descriptions. They are original teaching artifacts, not official statistics or a reproduction of the publisher's chart template.
Chart one: completed task counts
The bars below both start at the same zero baseline. Each filled block represents exactly one completed task. North's bar has fifteen blocks; South's has eighteen. The numbers at the right preserve the exact values. In both displays, N means North and S means South.
Completed tasks at the fictional snapshot
0 Completed tasks
N|███████████████ 15
S|██████████████████ 18
Data source: invented practice records in this guide. Each block is one task; both rows use the same scale. These are counts, not percentages or time values. The complete assigned totals remain in the input table.
The text description is: “At one fictional snapshot, North has fifteen completed tasks and South has eighteen. South's supplied completed count is three higher.” This description preserves the comparison without requiring the reader to distinguish colours or count blocks.
A chart title such as “South is more productive” would exceed the data. There is no duration or staffing denominator from which to calculate productivity. Use a title that identifies the actual measure.
Chart two: share of assigned tasks completed
For this separate chart, each filled block represents five percentage points. North's twelve blocks represent 60%; South's nine represent 45%. Both bars again start at zero and use the same scale. The counts beside them show the denominators.
Completed share at the fictional snapshot
0 Share completed
N|████████████ 60%
S|█████████ 45%
Data source: invented practice records in this guide. North is 15 of 25 assigned tasks; South is 18 of 40. Each block represents five percentage points, not five tasks.
The text description is: “North has completed 60% of its twenty-five assigned tasks, and South 45% of its forty. North's supplied share is fifteen percentage points higher.” This is a descriptive comparison of the fixture, not evidence of statistical significance or better management.
Keep the two scales separate. Copying the first chart's one-task-per-block legend onto the percentage chart would misdescribe it. Likewise, combining eighteen and sixty on an unlabelled axis would mix incompatible quantities.
Inspect a misleading baseline
Suppose a rejected count-chart design starts its numerical axis at fourteen instead of zero. The visible part of North's fifteen-count bar would have length one unit, while South's eighteen-count bar would have length four. That would make the latter visible segment four times as long.
The supplied counts are still fifteen and eighteen. Their ratio is 18 ÷ 15 = 1.2, not four. The extra apparent contrast arises from the changed baseline, not from four times as many completed tasks.
Repair this particular bar design by returning both bars to zero and keeping the exact labels. A note about the cropped axis would disclose it, but the zero-based display directly represents the magnitudes in this exercise.
The lesson concerns bars that encode magnitude through length from a baseline. It is not a rule that a line-chart detail view can never use a cropped axis. Evaluate the visual encoding and disclose the scale that applies.
Handle an absent value honestly
For a separate variation, suppose South's completed count is not supplied, while its assigned count remains forty. You cannot calculate its completed share from that input. The appropriate table entry is “not supplied,” and the rate remains unknown.
Do not display a zero-length South bar and call the count zero. Zero would be a known quantity; an absent value is missing evidence. Nor should you quietly reuse eighteen and 45% from the original fixture after declaring the completed count absent.
A clear variation caption could say: “North remains fifteen completed out of twenty-five; South's completed count is unavailable, so its completed share is not shown.” The changed-input display must stay separate from the baseline charts.
Check a denominator change
Another separate scenario retains South's eighteen completed tasks but changes its assigned total to thirty. Its rate becomes 18 ÷ 30 = 60%, matching North's 60%, while its completed count remains higher by three.
This is an invented denominator change, not a later observation or an improvement caused by an intervention. Retain the baseline assigned total of forty beside its original 45% rate. A reader should be able to identify which contract produced each number.
Identical rates also do not prove identical work or equal performance. They establish the same fraction for these supplied counts. The task differences and representative scope remain unknown.
Review the finished display as a reader
Read the chart title, legend, exact values and accompanying table together. Confirm that the labels refer to the same snapshot and that neither chart silently introduces a time trend. North and South are categories, so a connecting line would not represent a supplied chronological sequence here.
Check the display at the size where people will use it. Text that overlaps, tiny labels or an inaccessible source note can make a correct calculation hard to inspect. The code-style text charts above deliberately keep the lines short, but the actual page still needs layout verification.
You can ask a reader to state the count comparison and the rate comparison separately. That proposed check has not occurred in this exercise. Do not claim that the display improved understanding without actual responses.
Keep an honest visualization artifact
The complete practice packet contains the original data table, independently checked formulas, two text charts, their descriptions and the rejected-scale explanation. Keep the missing-value and changed-denominator scenarios labelled separately.
This artifact demonstrates a particular display and reasoning exercise. It does not establish proficiency with every visualization platform, a deployed dashboard, client approval or a salary advantage. If you later build the charts in software, record which tool you actually used and verify its output against the same input values.
For companion practice, the data-analysis guide addresses duplicate and missing records. The A/B-testing guide distinguishes percentage-point change from relative change and keeps descriptive results separate from inference. A useful visual makes those distinctions easier to inspect without enlarging the claim beyond its evidence.