Đại học Kinh tế Quốc dân – National Economics UniversityMICRO4ASIA
🏆 Arena
Sign inSign up
MICRO4ASIAThe six micro-credentials

Courses are delivered on the Limio learning platform. Limio

← Data Visualization in Business

You are viewing a public preview

Sign in to save your progress, do assignments, join discussions and get feedback.

Sign in
Module 1 — Overview of Data Visualization in Business

1.3 · Principles and Methods of Visualization Design

Print / Save as PDF

In this lesson

Visual encoding and design principles; choosing visual forms for the analytical objective, the data and the audience; use of color, position, size, labels and legends; clarity and accuracy. The slides present the three core requirements of a visualization, visual attributes, color practices, positioning and information density, and interaction techniques.

Learning outcomes

  • LO.1: Explain fundamental concepts of business data, data analysis, data visualization, chart types, dashboards and data storytelling.
  • LO.6: Select, build and refine charts that fit the analytical objective and the data.

The lecture slides for this lesson are below.

Study notes

By the end of this lesson you can:

  • Evaluate a chart against the three core requirements: clarity, accuracy and efficiency.
  • Name the main visual attributes and say which one suits which kind of data.
  • Choose a qualitative, sequential or diverging colour palette and assign background, dominant, accent and semantic colours.
  • Reduce clutter by adjusting positioning and information density.
  • Describe the main interaction techniques: filtering, dynamic queries, selecting and rearranging.

Key idea: Every design choice either helps the viewer read the data or gets in the way. Encode the most important comparison with position, use colour on purpose, and remove everything that does not carry information.

Three core requirements

RequirementMeaningEvaluation question
ClarityViewers understand the information quickly, without confusion or ambiguityCan viewers find the main message without guessing?
AccuracyThe chart faithfully represents the data, keeping relationships and proportionsAre values, scales and the size of differences shown correctly?
EfficiencyViewers absorb the information with minimal cognitive effortCan any element be removed, or the information shown more concisely?

Use these three questions as a checklist every time you finish a chart.

Visual attributes

Visual attributes (also called visual channels) are the properties of a mark that can carry data:

  • Position: where a mark sits in a coordinate system. It is the most accurate channel for comparing values, which is why bar, line and scatter charts work so well.
  • Shape: distinguishes groups or categories, for example circles for one segment and squares for another.
  • Size: shows magnitude of a quantitative measure, as in bubble charts. People judge area less precisely than position.
  • Colour: distinguishes categories, shows intensity or values, or highlights specific points.
  • Line width: distinguishes or emphasises lines.
  • Line type (solid, dashed, dotted): separates series or states, for example actual versus forecast.

Rule of thumb: put the comparison that matters most on position, and use colour, size and shape for secondary information.

Colour practices

Colour carries emotion and meaning. The lecture links red with urgency or danger (good for critical issues), blue with trust and professionalism (good for long-term performance) and yellow with positivity and attention (good for highlighting growth). Keep in mind:

  • Not everyone sees colour the same way; red-green colour blindness is common. Support colour with labels, patterns or symbols.
  • Avoid more than about five colours in one visualization, and avoid rainbow palettes.

Three palette types match three kinds of data:

  • Qualitative: different hues with no perceived order, for nominal categories such as Segment.
  • Sequential: one hue that changes in lightness or saturation in even steps, for ordered or quantitative values such as Sales.
  • Diverging: two sequential palettes back to back that meet at a neutral middle, for values that run from negative to positive, such as Profit.

Colours also have roles in a design: a background colour that is neutral and low in saturation; a dominant colour used consistently across visuals; an accent colour for focal points, alerts and KPIs, used sparingly; semantic colours with fixed meaning (red = loss, green = positive); and contrasting colours to separate data groups.

Positioning and information density

  • Remove elements that add nothing, such as a legend that repeats axis labels or an axis when every bar is labelled.
  • Reorganise visuals into a clean, structured layout.
  • Keep scales and labels consistent across visuals, so that the same colour or axis range means the same thing everywhere.
  • Reduce text: drop redundant axis titles and limit data labels to the values that matter.

Interaction

Interaction lets a visualization tell a dynamic story (over time, or as a sequence of steps played back) and lets users explore: zoom into details, create different views, change or filter values, and show connections between data. The lecture names four techniques:

  • Filtering: limit the displayed information with filter criteria.
  • Dynamic query: sliders or buttons that update the filtered data continuously.
  • Selecting: mouse click, hover with tooltip, drag or zoom.
  • Rearranging: move, sort or delete elements.

Worked example

A manager receives a bar chart of Profit by Sub-Category from the course dataset. It uses 13 rainbow colours, a legend, gridlines, unsorted bars and the title "Sheet 1". Redesign it step by step.

  1. Clarity. The question is "which sub-categories lose money?" Sort the bars by Profit and replace "Sheet 1" with a title that states what you see, in the form "[these sub-categories] lose money; the rest are profitable".
  2. Accuracy. Keep the axis starting at zero, so bars to the left of zero clearly show losses.
  3. Colour. Sub-Category is already shown by position and label, so colour per sub-category is redundant. Instead colour by Profit with a diverging palette (for example orange for negative, blue for positive, neutral at zero). Profit runs from negative to positive, which is exactly what diverging palettes are for.
  4. Efficiency. Remove the legend (the labels already name each bar), lighten or remove gridlines, and label only the negative bars.
  5. Interaction. Add a Category filter so the manager can focus on Furniture, Office Supplies or Technology.

The redesigned chart answers the question in a few seconds and is also readable by someone with red-green colour blindness, because position and labels carry the message even without colour.

Common mistakes

  • Using colour where position already works. Colour every bar differently and the viewer looks for a meaning that does not exist.
  • Sequential palette for signed values. A single-hue scale makes small losses and small profits look alike; use a diverging palette.
  • Rainbow and too many colours. More than about five colours cannot be told apart reliably.
  • Overusing the accent colour. If everything is highlighted, nothing is.

Summary

  • Check every chart for clarity, accuracy and efficiency.
  • Position is the most accurate visual attribute; use colour, size, shape and line style for secondary information.
  • Match the palette to the data: qualitative for categories, sequential for ordered values, diverging for values around a midpoint.
  • Remove clutter and keep scales and colours consistent across visuals.
  • Interaction (filter, dynamic query, select, rearrange) lets users explore beyond a single static view.

Flashcards · Principles and Methods of Visualization Design

9 words
WordMeaningExample
  • Clarity

    The design requirement that viewers can understand the main message quickly without guessing.

    A title that states the finding, not 'Sheet 1'.

  • Accuracy

    The requirement that a chart faithfully represents values, scales and proportions without misleading.

    Bar chart axes start at zero.

  • Efficiency

    The requirement that viewers absorb information with minimal effort, with no unnecessary elements.

    Remove a legend that repeats axis labels.

  • Visual attribute

    A property of a mark, such as position, shape, size, colour, line width or line type, that encodes data.

    Size of a bubble shows Sales.

  • Qualitative palette

    Distinct hues with no perceived order, used for nominal categories.

    Three hues for Consumer, Corporate, Home Office.

  • Sequential palette

    One hue in evenly increasing lightness or saturation, used for ordered or quantitative values.

    Light to dark blue for Sales by Province.

  • Diverging palette

    Two sequential palettes meeting at a neutral midpoint, for values ranging from negative to positive.

    Orange for losses, blue for profit.

  • Accent colour

    A colour reserved for focal points, alerts or KPIs; it loses its effect if overused.

    One highlighted bar, all others grey.

  • Dynamic query

    An interaction where sliders or buttons continuously update which data are shown.

    A date slider that redraws the sales trend.