Tables, Graphs, Charts & Diagrams
Open in the Padhaiko appVisual Communication in Technical Engineering
Visual aids are fundamental to technical communication. Technical documents and engineering presentations rarely rely on text alone; empirical data, system architectures, physical mechanisms, and process algorithms are comprehended far more rapidly and accurately when presented visually.
Graphics serve five decisive functions in technical documentation:
- Demonstrating Logical and Numerical Relationships: Presenting numerical trends, correlations, and structural hierarchies clearly.
- Communicating Spatial Information: Illustrating physical shapes, dimensions, and component locations that would require pages of ambiguous verbal description.
- Clarifying Steps in a Process: Mapping sequential execution flows, decision forks, and procedural instructions.
- Saving Document Space and Reader Time: Condensing complex multivariate datasets into concise visual matrices.
- Aiding International Comprehension: Bridging linguistic barriers for multilingual engineering teams.
Choosing the Right Visual Aid
Selecting an inappropriate graphic can obscure critical data and mislead readers. The choice of visual format is dictated by the type of data and the communicative objective.
| Communicative Objective | Recommended Visual Aid | Key Advantages and Typical Use Case |
|---|---|---|
| Lookup of exact numerical values or multi-parameter comparison | Table | Displays exact numerical readings across multiple variables (e.g., transistor characteristics, server specifications). |
| Continuous trends over time or correlation between two variables | Line Graph | Illustrates rates of change, continuous functions, and trends over time (e.g., turbine efficiency vs. flow rate). |
| Comparison of discrete quantities across categories | Bar Chart | Highlights differences in magnitude among distinct categories (e.g., quarterly power generation by hydro plant). |
| Proportions of a single whole (100% total) | Pie Chart | Displays relative percentage breakdown among a small number of components (e.g., energy source distribution in national grid). |
| Sequential logic, algorithms, and decision workflows | Flowchart | Shows procedural sequence, conditions, and decision branches (e.g., compiler parsing phases, emergency shutdown protocol). |
| System hierarchy, organizational reporting, and modular divisions | Organizational Chart / Block Diagram | Depicts architectural modules, subsystem boundaries, and data interconnects (e.g., microprocessor bus architecture). |
| Internal structural mechanisms and spatial arrangements | Cutaway / Cross-Section Drawing | Reveals interior parts and functional assemblies otherwise hidden inside an enclosure (e.g., Pelton turbine nozzle). |
| Assembly, disassembly, and maintenance procedures | Exploded-View Diagram | Shows separated components aligned along assembly axes to illustrate how parts fit together (e.g., motor chassis assembly). |
Standards for Labelling, Numbering, and Referencing Visuals
To integrate smoothly with technical prose, visual aids must adhere to strict academic and professional documentation conventions.
Captioning and Referencing Standards
- Distinct Numbering Sequences:
- Tables and figures maintain independent, sequential numbering systems throughout a document: Table 1, Table 2, etc., and Figure 1, Figure 2, etc. (or chapter-prefixed: Table 3.1, Figure 4.2).
- All non-table graphics (graphs, charts, diagrams, drawings, photographs, maps) are classified under the universal umbrella term Figure.
- Placement of Titles and Captions:
- Table Caption: Placed ABOVE the table, left-aligned or centered, in bold text followed by an informative title (Table 2: Dynamic Power Consumption Across CPU Clock Frequencies).
- Figure Caption: Placed BELOW the graphic, left-aligned or centered, followed by an explanatory title and legend (Figure 4: Functional Block Diagram of the 8086 Bus Interface Unit).
- In-Text Cross-Referencing (The Lead-In):
- Every visual aid must be formally introduced and discussed in the preceding text before it appears on the page. Never drop a graphic into a document without an explicit in-text reference.
- Effective lead-in: "As summarized in Table 3, thermal dissipation rises exponentially once clock frequencies exceed 3.2 GHz."
- Avoid empty mechanical references: "Table 3 shows data." Always explain what technical conclusion the reader should draw from the graphic.
- Source Attribution:
- If a graphic or its underlying data is borrowed or adapted from an external publication, provide full attribution directly beneath the graphic (e.g., Source: Adapted from NEA Annual Progress Report 2082/83).
6.1 Tables
A table is an orderly arrangement of data into horizontal rows and vertical columns. It is the most effective format for presenting detailed, exact numerical figures, multivariate comparisons, and structured textual specifications.
Anatomy of a Formal Engineering Table
| Component | Definition and Formatting Rules |
|---|---|
| Table Number & Title | Centered or flush left above the table. Clear, specific, and self-contained (e.g., Table 1: Electrical Specifications of STM32F401 Microcontroller). |
| Box Head (Column Heads) | The topmost row containing headers that identify what data appears in each column below. Must include explicit units of measurement in parentheses (e.g., Current (mA), Latency (ns)). |
| Stub (Row Heads) | The leftmost vertical column identifying the items, categories, or test conditions for each horizontal row. |
| Data Cells (The Body) | The intersections of rows and columns containing the specific numbers, text, or status indicators. |
| Footnotes & Source Notes | Placed directly beneath the table to define abbreviations, explain anomalous readings, or cite external data origins. |
Design and Formatting Guidelines for Tables
- Alignment: Right-align all numeric data to align decimal points vertically. Left-align text descriptions; center short status codes, dates, or symbols.
- Minimal Gridlines: Avoid heavy, dense gridlines that produce visual clutter. Use horizontal rules to demarcate the top and bottom of the table and separate the box head from the body; omit vertical internal rules unless required to separate distinct grouped variables.
- Zero and Missing Values: Never leave a data cell blank. An empty cell suggests a typographical omission. Use N/A (not applicable), a dash (—), or an explicit 0.0 to indicate measured zero.
- Logical Ordering: Arrange rows systematically (alphabetically, by decreasing magnitude, chronologically, or by functional hierarchy) rather than randomly.
6.2 Graphs
A graph maps numerical data onto a two-dimensional coordinate system, transforming abstract numbers into spatial geometric curves. Graphs excel at illustrating trends over time, rates of change, and empirical relationships between independent and dependent variables.
Line Graphs
Line Graph
A line graph plots data points on a Cartesian grid (X and Y axes) connected by continuous line segments. The horizontal axis (X-axis, abscissa) represents the independent variable (the variable controlled or measured sequentially, such as time, temperature, or distance), while the vertical axis (Y-axis, ordinate) represents the dependent variable (the resulting measurement, such as voltage, efficiency, or packet loss).
Best Practices for Engineering Line Graphs
- Axis Calibration and Scale: Choose scales that accurately portray data without exaggeration. Marking tick intervals clearly with legible numerical values.
- Multiple Data Curves: When plotting several curves on a single graph, restrict the total to no more than 3 to 5 lines to prevent clutter.
- Curve Differentiation: Do not rely solely on color to distinguish lines, as documents are frequently photocopied or printed in monochrome. Differentiate curves using distinct line styles (solid, dashed, dotted) and distinct geometric data markers (circles, squares, triangles).
- Direct Curve Labelling: Place labels directly adjacent to the lines whenever possible rather than isolating them in a distant legend box.
Scatter Plots
A scatter plot displays individual bivariate data points without connecting lines. It is used in empirical engineering research to explore the strength and direction of correlation between two physical phenomena (e.g., concrete compressive strength vs. water-cement ratio). A regression line (trendline) is typically fitted to indicate the underlying mathematical function.
6.3 Charts
While graphs plot continuous mathematical functions across coordinate axes, charts represent discrete categorical relationships, quantities, hierarchical divisions, or algorithmic workflows.
Bar Charts
Bar Chart
A bar chart uses horizontal bars or vertical columns of varying length to compare the magnitudes of discrete categories. The length of each bar is directly proportional to the value it represents.
Variations of Bar Charts
- Vertical Column Chart: Standard format used when categories are chronological (e.g., annual energy generation from 2020 to 2026).
- Horizontal Bar Chart: Preferred when category names are long (e.g., comparing 12 municipality names), as horizontal labels can be read naturally from left to right without diagonal tilting.
- Grouped (Clustered) Bar Chart: Compares multiple sub-categories across main groupings (e.g., comparing peak power demand vs. actual generation across four seasons). Limit to 2–3 bars per cluster.
- Segmented (Stacked) Bar Chart: Divides each bar into proportional components to show both total magnitude and internal percentage composition. Effective only when the relative contribution of each sub-element is significant.
The Mandatory Zero-Baseline Rule
Bar charts convey magnitude through the visual area and length of the bar. The numerical scale of a bar chart must always begin at zero. Truncating the baseline (e.g., starting an axis at 90 instead of 0) artificially magnifies microscopic differences into dramatic visual chasms, constituting a severe violation of technical ethics.
Pie Charts
Pie Chart
A pie chart is a circular graphic divided into radial sectors, where each sector represents a percentage proportion of a single, unified whole (100%).
Construction Rules for Pie Charts
- Strict 100% Total: All slices combined must equal exactly 100%. Never use a pie chart to plot non-additive metrics or multi-choice survey selections.
- Limit Number of Slices: Restrict slices to between 3 and 7. A pie chart with 15 thin slivers is unreadable; combine minor fractions into a single grouped category titled "Other".
- Logical Sector Ordering: By convention, place the largest slice starting at the 12 o'clock vertical position and arrange subsequent slices in descending order clockwise.
- Direct Labelling: Place category titles and exact numerical percentages directly on or adjacent to each slice rather than using a separate, detached color key.
Flowcharts and Process Charts
Flowchart
A flowchart is a standardized diagram that maps the sequential steps of an algorithm, manufacturing workflow, or decision-making protocol using standardized geometric symbols connected by directional arrows.
| Geometric Symbol | Traditional Technical Name | Operational Function in Workflow |
|---|---|---|
| Oval / Rounded Rectangle | Terminal Symbol | Indicates the definitive Start or End / Exit point of a process. |
| Rectangle | Process Symbol | Represents a single operational task, computation, or algorithmic execution step (e.g., Compute FFT coefficients). |
| Diamond | Decision Symbol | Represents a conditional branch or test condition with two or more exiting paths labeled Yes/No or True/False (e.g., Is Buffer Full?). |
| Parallelogram | Input / Output Symbol | Represents manual or automated data entry or output generation (e.g., Read sensor voltage on Pin A0). |
| Circle / Small Shape | Connector Symbol | Links disparate sections of a flowchart across page breaks or complex loops without crossing lines. |
| Directional Arrow | Flowline | Shows the exact direction of programmatic or physical sequence. |
Gantt Charts and Organizational Charts
- Gantt Chart: A horizontal timeline chart mapping project tasks along a calendar axis. It illustrates task durations, dependencies, concurrency, and milestone deadlines across engineering sprints.
- Organizational Chart: A tree-structured hierarchical diagram mapping administrative authority, reporting lines, and division of responsibility within an enterprise.
6.4 Diagrams
A diagram is a specialized, non-photographic line drawing that represents physical structures, operational principles, or electrical connections. Unlike realistic photographs that capture surface clutter and shadows, diagrams deliberately strip away extraneous visual noise to highlight functional architecture and component interactions.
Types of Engineering Diagrams
Core Engineering Diagrams
- Schematic Diagrams: Abstract representations of functional systems using standardized symbolic glyphs rather than realistic physical drawings.
- Circuit Schematics: Use international IEEE/IEC symbols for resistors, capacitors, operational amplifiers, and logic gates to show electrical connectivity.
- Piping and Instrumentation Diagrams (P&ID): Show process piping, pumps, valves, and telemetry sensors in power plants and chemical refineries.
- Block Diagrams: High-level conceptual diagrams where functional subsystems are represented as simple labeled geometric boxes connected by directional data or power buses. Essential for explaining microprocessor architectures, network topologies, and control loops.
- Cutaway and Cross-Section Drawings: Drawings that simulate cutting away an external casing or slicing through a mechanical assembly to reveal internal functional elements (such as bearings, impellers, or combustion chambers) in their precise spatial relationships.
- Exploded-View Diagrams: Technical illustrations showing an assembly separated into its constituent parts, with each part floating along an assembly centerline axis. Widely used in user manuals, parts catalogs, and maintenance guides to show assembly sequence and part numbers.
| Aspect | Realistic Photographs | Technical Diagrams |
|---|---|---|
| Visual Content | Captures actual physical surface textures, real-world colors, shadows, and environment | Abstracts away surface clutter, shadows, and dirt to display functional mechanisms |
| Internal Structures | Cannot reveal interior components concealed behind opaque metal or plastic enclosures | Readily reveals internal assemblies via cutaway, cross-section, and exploded perspectives |
| Clarity of Function | Background reflections and wiring clutter can obscure circuit connectivity | Standardized symbolic glyphs make electrical logic and signal flow unambiguously clear |
| Reproduction | Degrades in low-cost photocopying and printing | High-contrast vector linework remains crisp across all reproduction media |
Ethics in Visual Representation: Avoiding Visual Distortion
Technical communicators have an ethical obligation to portray data honestly. Graphs and charts can easily be manipulated to mislead unsuspecting readers.
Chartjunk and Visual Distortion
Chartjunk is a term coined by Edward Tufte to describe visual ornamentation, unnecessary 3D effects, moiré patterns, and artificial decorations that clutter a graphic without adding information.
Common visual distortions that violate engineering ethics include:
- Truncated Vertical Axes (Suppressed Zero): Cutting off the bottom of a bar chart axis to artificially magnify trivial percentage fluctuations into massive visual differences.
- Artificial 3D Perspective Distortions: Applying tilted 3D perspectives to pie charts or bar charts. Tilting a pie chart distorts angles: the sector nearest the viewer appears substantially larger than an identical or larger sector at the back.
- Non-Linear Axis Scaling: Distorting tick intervals unevenly along an axis to flatten or exaggerate slopes.
- Scale Inflation: Altering width and height simultaneously when scaling pictorial bar symbols (pictograms), so that doubling a value quadruples the visual surface area.
Worked Examples
1
Evaluating Narrative Weaknesses: Presenting multidimensional numerical specifications in running prose forces the reader to re-read sentences repeatedly to locate, extract, and compare parameters across platforms. Memory sizes, voltages, and costs cannot be scanned rapidly.
Designing the Table Structure:
- Create an informative table title placed above the table.
- Define a Box Head with explicit parameter units: Operating Voltage (V), Clock Speed (MHz), Active Current (mA), Flash Storage (KB), SRAM (KB), Unit Price (NPR).
- Right-align all numeric quantities to preserve decimal alignment.
- Include an explanatory footnote regarding integrated wireless hardware.
Formatted Technical Table:
Table 1: Comparative Hardware Specifications and Procurement Costs of Evaluated Microcontrollers
| Microcontroller Platform | Operating Voltage (V) | Max Clock Speed (MHz) | Active Current (mA) | Flash Memory (KB) | Internal SRAM (KB) | Unit Price (NPR) |
|---|---|---|---|---|---|---|
| ATmega328P | 5.0 | 16 | 9.2 | 32 | 2 | 450.00 |
| STM32F401 | 3.3 | 84 | 18.5 | 256 | 64 | 1,250.00 |
| ESP32-WROOM-32* | 3.3 | 240 | 80.0 | 4,096 | 520 | 850.00 |
Note: Active current consumption for the ESP32 includes active 2.4 GHz Wi-Fi radio telemetry state.
2
Analyzing Ethical Violations and Visual Distortion:
- Suppressed Zero Baseline: By truncating the baseline and beginning the vertical axis at 95 ms instead of 0 ms, the visual bar height is decoupled from mathematical magnitude.
- Deceptive Visual Ratio: The actual performance gain is approximately 5.8% (104 - 98 = 6 ms out of 104 ms). However, the chart visually communicates an apparent 80% to 90% reduction, deceiving viewers who scan visual bar heights without scrutinizing tick labels.
- Violation of Engineering Integrity: Presenting data in a way that creates a false cognitive impression violates the IEEE Code of Ethics and foundational principles of honest technical communication.
Prescribing Corrective Actions:
- The vertical axis of the bar chart must start at 0 ms.
- When plotted from zero, the bars will show their true relative heights, honestly depicting the modest 5.8% latency reduction.
- If the engineer wishes to highlight the 6 ms reduction specifically, they should use a table or plot a line graph depicting latency over successive optimization iterations, accompanied by explicit percentage annotations.
Diagnostic Summary and Corrective Design: The original chart commits the visual ethics violation of a truncated vertical axis (suppressed zero baseline). While the physical reduction in execution time is only 5.77% (a 6 ms decrease), starting the axis at 95 ms makes the "Before" bar look 600% taller than the "After" bar, visually deceiving the reader.
Corrective Remedy:
- Re-anchor the vertical axis at an honest baseline of 0 ms, extending to 120 ms.
- Label both bars with their exact execution values (104 ms and 98 ms).
- Annotate the graph honestly: "Measured latency reduction: 5.8% (6 ms)".
Exercises
Exercises — Visual Aids in Communication
State the formal captioning, numbering, and in-text lead-in conventions for tables versus figures in an engineering report. Why must tables and figures maintain independent numbering sequences?
Analyzing Placement and Numbering Rules:
- Tables: Caption is placed above the table (e.g., Table 1: Title).
- Figures: Caption is placed below the visual (e.g., Figure 1: Title). Figures encompass all non-table visuals including graphs, charts, diagrams, drawings, and photos.
- Independent Sequences: Tables and figures maintain independent numbering (Table 1, Table 2 alongside Figure 1, Figure 2) to simplify document indexing, prevent confusion in tables of contents, and allow readers to differentiate tabular numerical matrices from graphical illustrations immediately.
Evaluating the In-Text Lead-In: Every visual must be explicitly referenced in the narrative text before it appears ("As illustrated in Figure 3..."). The text must explain the key insight or takeaway, not just point to its existence.
Table titles are placed above the table; figure captions are placed below the graphic. They maintain independent numbering sequences (e.g., Table 1, Table 2; Figure 1, Figure 2) to maintain distinct visual indexing in formal lists of tables and figures. Furthermore, every visual must have an in-text lead-in introducing and explaining its key findings in the prose before the visual appears on the page.
Compare the specific communicative functions of a Line Graph, a Bar Chart, and a Pie Chart. For what type of data is each visual uniquely suited?
Comparing Visual Applications:
- Line Graph: Suited for continuous quantitative data showing trends over time or continuous functional relationships between an independent variable (X) and dependent variable (Y).
- Bar Chart: Suited for discrete categorical comparisons, highlighting relative differences in magnitude across distinct entities.
- Pie Chart: Suited exclusively for showing proportional percentage divisions of a single unified whole (100%) across a limited number of categories (3 to 7).
A Line Graph is best for continuous numerical variables to display rates of change, trends over time, or mathematical correlations. A Bar Chart is ideal for discrete categorical comparisons of magnitude across distinct groups. A Pie Chart is suited strictly for displaying the percentage share of components comprising a single 100% whole.
Why must the scale of a bar chart always begin at zero? Explain how suppressing the zero baseline violates ethical technical communication.
Understanding Visual Perception of Bar Charts: Viewers judge the magnitude of bar chart values by the physical length and surface area of the bars.
The Mechanism of Distortion: When the zero baseline is suppressed (e.g., starting an axis at 90 instead of 0), a value of 92 has a bar height of 2 units, while a value of 98 has a bar height of 8 units. Visually, the second bar appears four times (400%) larger than the first, even though the actual mathematical difference is only 6.5%. This creates an intentional or negligent visual deception, violating the fundamental ethical standard of technical honesty.
The numerical scale of a bar chart must always start at zero because readers perceive magnitude through the relative length and visual area of the bars. Suppressing the zero baseline artificially magnifies minute numerical variations into dramatic visual disparities, misrepresenting empirical facts and violating ethical engineering standards.
Describe the standard ANSI/ISO flowchart symbols for Terminal, Process, Decision, and Input/Output operations. Explain how decision symbols control branch execution in algorithm diagrams.
Identifying Standard Symbols:
- Terminal (Oval / Rounded Rectangle): Marks the start or end of the process.
- Process (Rectangle): Represents a discrete computational step or action.
- Decision (Diamond): Represents a conditional test branching into alternate execution paths.
- Input/Output (Parallelogram): Represents data entry or result output.
Explaining Decision Branch Control: A decision diamond evaluates a logical condition (e.g., Is Temperature > 80°C?). It has one entry path and two or more labeled exit flowlines (typically Yes/No or True/False). Depending on the evaluation of the condition, execution branches along the corresponding labeled path, enabling loops, error handling, and conditional routing.
The standard symbols are: Oval for Terminal (start/end), Rectangle for Process (calculation/action), Diamond for Decision (conditional test), and Parallelogram for Input/Output (data transfer). The decision diamond controls branching by evaluating a logical condition and directing program flow along two or more mutually exclusive paths labeled with outcomes (e.g., Yes/No), thereby governing loops and conditional execution.
Contrast an exploded-view diagram with a cutaway drawing. In what types of engineering documentation is each format indispensable?
Defining the Two Diagram Types:
- An exploded-view diagram separates all components of a mechanical assembly, drawing them separated along an axis of assembly to show how individual parts fit together.
- A cutaway drawing slices away a portion of the exterior enclosure to reveal the internal spatial relationships of components within an operating assembly.
Identifying Appropriate Documentation:
- Exploded-view diagrams are indispensable in assembly instructions, parts replacement catalogs, and maintenance repair manuals.
- Cutaway drawings are indispensable in technical sales brochures, training manuals, and engineering textbooks to illustrate the operational mechanics of pumps, engines, valves, and sealed assemblies.
An exploded-view diagram illustrates components separated along assembly axes to demonstrate how parts interlock; it is indispensable in maintenance manuals, assembly guides, and parts catalogs. A cutaway drawing cuts away part of an exterior shell to reveal internal operational mechanisms; it is essential in training manuals and system descriptions explaining the internal mechanics of enclosed machines.
What is "chartjunk"? Provide three specific examples of visual ornamentation that degrade the technical clarity of engineering graphics.
Defining Chartjunk: Coined by Edward Tufte, chartjunk refers to unnecessary visual elements, ornamentation, and graphic flourishes that clutter a visual aid without contributing substantive data or meaning.
Providing Examples:
- Unnecessary 3D effects: Adding simulated depth (extrusion) to 2D bar or pie charts, which obscures data point alignment and distorts angles.
- Dense decorative gridlines: Heavy, dark crosshatching and background grids that overpower the actual data curves.
- Clip art and decorative illustrations: Irrelevant cartoon figures or decorative drawings (e.g., clip-art computers) placed inside charts, distracting the reader's focus from data analysis.
Chartjunk refers to decorative visual ornamentation that clutters a graphic without conveying substantive technical information. Three common examples are: (1) Unnecessary 3D extrusion on 2D bar or pie charts that distorts data angles, (2) heavy, dark background gridlines that visually overpower data curves, and (3) irrelevant clip art or decorative icons that distract from analytical data evaluation.
Explain why a pie chart should never be used to represent non-additive categories or data whose sum exceeds 100%. What visual alternative should be selected instead?
Analyzing the Mathematical Nature of Pie Charts: A pie chart is geometrically constructed around the principle that the complete circle represents a unified whole (360 degrees = 100%). Each slice is an angular proportion of that single entity.
Identifying Failure Cases and Alternatives: If categories are non-additive (such as multiple-choice responses where a respondent can select more than one option) or represent independent percentages, their sum will not equal 100%. Plotting such data on a pie chart is a mathematical and visual fallacy. The correct alternative is a horizontal or vertical bar chart, where each bar independently represents its category's percentage without implying that they combine to form a closed whole.
A pie chart geometrically represents the division of a single unified whole where the sum of all sectors must equal exactly 100%. If categories are non-additive or allow multiple responses, summing them is mathematically invalid and visually misleading. A horizontal or vertical bar chart must be used instead to display independent category percentages accurately.