Monthly Archives: December 2014

Principles of Analytic Graphics

Principle 1: Show comparisons: show evidence for a hypothesis is always relative to another hypothesis.

Principle 2: Show relationships: The relationships can be causality, mechanism, explanation, systematic structure. This is the purpose of an analytic graphics.

Principle 3: Show multivariate data: The real world is always complex and multivariate. It’s more likely to reflect the true picture using multivariate data. Otherwise, we assume all other factors not included are the same (or homogeneous) for each group displayed.

Principle 4: Integrate multiple modes of evidence: Each mode of evidence explains one perspective of the relationships. Integration of multiple modes can then better disclose the full spectrum of the relationships.

Principle 5: Describe and document the evidence: A data graphic should tell a complete story that is credible. It helps develop credibility to document a graph with appropriate labels, scales, sources, etc.

Principle 6: Content is king. 

adapted from course note of ‘Exploratory Data Analysis’ (John Hopkins Data Science Specialization course).