Data Visualization Flashcards
What are visualizations?
combine and integrate visual marks into more complex structural forms to encode information
What do visualizations help do?
- illustrate relations
- discover trends, patterns, and outliers
- get and keep the attention of recipients
- support remembrance and recall
- facilitate learning
- motivate people and establish a mutual story and initiate actions by illustrating options to act
What are the 9 Gestalt Principles?
- proximity
- similarity
- enclosure
- connection
- continuity
- symmetry
- figure and ground
- closure
- common fate
Gestalt Principle
What is proximity?
how close elements are to one another – similar things should be close to each other
Gestalt Principle
What is similarity?
people tend to see things that physically resemble each other as part of the same object
Gestalt Principle
What is enclosure?
we group elements that are in the same closed region
Gestalt Principle
What is connection?
grouping effect – we perceive elements as connected to each other thanks to colors, lines, frames, or other shapes
Gestalt Principle
What is continuity?
objects that create a continuous pattern or are seen as being connected appear to be grouped together
Gestalt Principle
What is symmetry?
elements that are symmetrical tend to be perceived as a unified group
Gestalt Principle
What is figure and ground?
your brain distinguishes foreground and background
Gestalt Principle
What is closure?
our eyes tend to add any missing pieces of a familiar shape
Gestalt Principle
What is common fate?
people will group together things that point to or are moving in the same direction
What are the 4 data presentation principles?
- understand the purpose – exploratory vs. explanatory
- understand what is to be communicated – most important message you are trying to convey
- choose appropriate visual representation
- quality of presentation – put more emphasis on (from higher to lower): location, size, colour
What are the 5 principles for creating infographics?
- simplicity
- consistency
- visibility
- navigability
- suitablity
Principles for Creating Infographics
What is simplicity?
minimal text
clear message
avoid clutter
Principles for Creating Infographics
What is consistency?
layout and design elements should be consistent
- 2-3 font sizes
- color scheme
Principles for Creating Infographics
What is visibility?
appropriate font sizes
colors that contrast
Principles for Creating Infographics
What is navigability?
(structure)
clear order to follow
use scale/proportion to emphasize key points/headings
Principles for Creating Infographics
What is suitability?
right data for message
right graphic for the message
right metaphors for the audience
Common Patterns – Visualization vs. Infographic
visualization
- exploratory
- lots of data
- informs
- little text
infographic:
- explanatory
- little data
- persuades
- more text
Scales for Comparing Values – More Accurate to Less Accurate
position common scale position non-aligned scale length direction angle area volume curvature shading color saturation
What are the visualization formats for comparing categories?
bar chart:
- bar chart
- stacked bar chart
- grouped bar chart
pie chart
What scale does bar chart use?
position common scale
What are bar charts good for?
- comparisons
- bar chart’s discrete data is categorical data, therefore answers the question of “how many” in each category
What is stacked bar chart good for?
What are limitations?
proportions, parts to a whole, comparisons
limitations:
- more segments per bar = harder to read
- hard to compare segments to each other, as they’re not aligned on a common baseline
What are grouped bar charts good for?
What are limitations?
distribution, relationships, comparisons
limitation:
- harder to read the more bars you have in one group
What are pie charts used for?
comparison, parts to a whole, proportions
What are limitations of pie charts?
- can only show few values
- takes up more space than alternatives
- not good for accurate comparisons
What scale does pie chart use?
area
When is it good to use pie chart?
very rare occasions – ie. in interactive situations, looking to see if exactly two things are equal
What is an area chart?
line graphs but with area below the line filled in
used to show trends rather than convey specific values, distribution
What is a stacked area chart similar to?
bar chart, but the focus is trend over time
good for: comparisons, trend over time
What are visualization formats for multiple attributes?
- bubble chart
- parallel coordinates
- radar chart
What are parallel coordinate charts?
- common way of representing multivariate numerical data
- each variable is given its own axis
What are parallel coordinate charts good for?
What are limitations?
good for: comparisons, relationships
limitations: over-cluttered, use interaction to address this
What are radar charts?
- used for comparing multiple quantitative variables
- useful for seeing which variables are scoring high or low
- relationships
What are limitations of radar charts?
- cluttered when we have multiple polygons or variables
- not good for comparing values across each variable
What are bubble charts?
- multivariable graph – cross between scatter plot and area chart
- uses Cartesian coordinate system to plot points on grid
- X and Y are separate variables
- uses size to represent another variable
- can use colour
- 4 variables
What are bubble charts good for?
What are limitations?
good for: comparisons, trends over time, distribution, proportions, relationships
limitations: area for comparison is not good
What are visualization formats for relationships and hierarchies?
- tree map
- scatterplot
- tree diagram
- network graph
What is a tree map?
- alternative way to representing hierarchical structure
- displays quantities for each category
What is a scatterplot good for?
relationships (ie. correlation), paired numerical data
can be 3D
What are tree diagrams?
often used to show relations and descent
What are network graphs?
shows how things are interconnected through use of nodes/vertices, and link lines to represent their connections and help illuminate the type of relationships between a group of entities
What are limitations of network graphs?
- limited data capacity
- starts to become hard to read when there are too many nodes (resemble ‘hairballs’)
What are visualization formats for visualizing spatial data?
- map with embedded charts
- chloropleth map
What is a chloropleth map?
display divided geographical areas or regions that are colored, shaded or patterned in relation to a data variable
provides a way to visualize values over a geographical area, which can show variation or patterns across the displayed location
requires legend
What are chloropleth maps food for?
What are limitations?
good for: comparisons, location
downside:
- use of colour – can’t accurately read/compare values from map
- larger regions appear more emphasized than smaller ones, so viewer’s perception of shaded values are affected
What are elaborate visualization formats?
- word cloud
- stream graph
- alluvial or sankey diagram
What is a word cloud?
displays how frequently words appear in a given body of text, by making size of each word proportional to its frequency
What is a word cloud good for?
What are limitations?
good for: proportions, text analysis, distribution
limitations:
- long words are emphasized over short words
- not great for analytical accuracy – used more for aesthetic purposes
What are limitations of static visualizations?
- difficult to represent large amounts of data
- inability to properly support the question and answer process involved in data analysis
What do interactions allow us to do?
- control flow of data
- be active participants in the analysis of data
- adjust features of the tool to suit the user’s needs
What are ethical considerations of visualizations?
- not fact
- accuracy depends on quality of data used, representation of data, objective of the creator
What is bad quality data?
- incorrect data
- missing data
- insufficient data
- GIGO
How can data representation be misleading?
- truncated graphs
- exaggerated scaling
- improper extraction
- misusing metaphors (ie. is green good or bad, is red good or bad)
- flawed representations