1. Principles of visual analytics
Let's talk about how to make our dashboard informative and actionable.
2. Visual analytics
Our theoretical foundation lies in two fields.
Information design is about how to make data accessible for specific audiences in specific situations to meet defined objectives. And data visualization provides visual methods of displaying data.
We use data to present evidence.
3. Visual analytics
Information design simplifies complex data facilitating our reasoning process.
4. Visual analytics
Visual analytics we enable human-information discourse where users interact with data to gain insights.
5. Fundamental principles of visual analytics
The principles for visual analytics were introduced by Edward Tufte in 2006.
6. Fundamental principles of visual analytics
These principles are universal and serve as a guide for how to accurately portray data using visual elements.
7. Fundamental principles of visual analytics
The meaningful comparison shows contrasts and differences between variables.
8. Fundamental principles of visual analytics
The causal structure shows the relationship and how one variable influences another.
9. Fundamental principles of visual analytics
The multivariate links principle demonstrates how complex data can be combined for easier interpretation.
10. Fundamental principles of visual analytics
Integration incorporates various models of information.
11. Fundamental principles of visual analytics
Documentation includes elements such as data origin and description for credibility.
12. Fundamental principles of visual analytics
And content describes the importance of data relevance and context.
13. Comparison and contrast
Showing comparisons, contrast, and differences is the basis of any scientific exploration. This is why we often ask the question, "Compared to what?" when presenting data.
But not all comparisons are meaningful or useful.
To make intelligent and appropriate comparisons, we can use objects, such as:
time,
different components,
or different groups.
14. Causality and relationship
We use the concept of causality and systematic structure to explain and connect relationships.
We can visualize these relationships in various ways, for example,
using a cause-and-effect diagram or fishbone to sort out ideas,
15. Causality and relationship
a scatter plot to display associations between variables,
16. Causality and relationship
and a causal graph to represent assumptions and relationships.
17. Causality and relationship
A cause-and-effect flowchart is a simple way to show causation without any statistical components.
18. Causality and relationship
Granger causality is a probabilistic concept of causality that uses the fact that causes must precede their effects in time.
19. Multiple variables
Real-world data is often complex, and it is common to work with more than two variables or dimensions. But we typically represent visual data using 2-dimensional graphs. To do so effectively, we have two main approaches:
Grouping: this involves adding visual characteristics such as color, shape, size, line type, or transparency to show the data for multiple groups on a single graph.
20. Multiple variables
Faceting: this approach creates a figure with multiple subplots, each showing a different subset of the data, but all sharing the same axes.
21. Integration of evidence
To fully comprehend data, it is essential to present it in a context, such as text, tables, images, or diagrams.
What matters is the evidence, not a particular mode of evidence.
So by integrating text and tables within the same graph, you can provide stronger visual evidence. In this example, you see the integration of statistical results into a graph.
22. Documentation and content
A good data graphic should be able to tell a complete story on its own.
Data graphics should provide information about data, visualization process, and data description.
To ensure your graphic is effective, be sure to answer key questions such as
where the data came from,
how it was processed and cleaned,
and whether a legend, labels, and graph title are included.
The success of analytical presentations depends on quality, relevance, and integrity.
To improve your content,
start with a strong question,
develop a sound approach,
and present all necessary information to answer that question.
23. Let's practice!
It is time to put these principles into practice!