LANGUAGE IN PERSONALITY SPACE
This visualization explores the relationships between users' language, personality, age, and gender. The data comes from over 70,000 Facebook users who completed personality tests and allowed their status messages to be analyzed.
Steps for Use:
- Using the first two tabs to the left, select a personality spectrum for the horizontal axis of the plot and a personality spectrum for the vertical axis of the plot. Here, the spectrum of Agreeableness has been selected for the horizontal axis and the spectrum of Extraversion has been selected for the vertical axis.
- Click the Features tab to determine if you would like to plot either words or topics (topics are relevant groups of words).
- Click the Color tab to determine if the plot should show word or phrase differences by age or by gender.
- Hover over a point on the plot to see where the words or phrases shown fall on the personality spectrums as well as their specific correlations with the given personality traits.
Each dot is a language topic.
Topics are groups of words that tend appear close together in language. Hover over a dot to see the topic's words and other details. The words that form each topic are chosen automatically by a computer algorithm (learn more about that here).
Topic use correlates with personality.
The vertical and horizontal position of each topic is determined by two personality dimensions (learn more about these dimensions here). The Horizontal and Vertical axis buttons change the axes dimensions that define this space.
Topic use correlates with gender and age.
The color of each topic is determined by its correlation with gender or with age — toggle this with the Color button at the top. A high gender correlation, for example, means that either men or women tend to use this topic more. Likewise, correlations with age mean that either younger or older users use this topic more. A small correlations (near zero) means that the topic is used equally across groups.
Language in Personality Space was developed largely by Gregory J. Park, now an independent Data Scientist and Web Visualization specialist. You can see more of his work here
