Big Five Personality Test

psychology | data visualization | machine learning

roles: machine learning researcher | data analyst

Motivation

I was curious to see how I could tell the personality type of millions of people simply from them answering a survey. I used machine learning to reduce the number of attributes when the relationships between them were not that obvious, like something as variable as a person’s openness or extraversion. It was beautiful to see a person’s psyche quantitatively and reflected in OCEAN (Openness, Conscientiousness, Extraversion, Agreeableness, Neuroticism) scores based on social psychology’s Big-Five Factor Markers from the International Personality Item Pool (Goldberg, 1992). Where for instance, a non-messy and prepared person would be scored as highly conscientious.

Description

Coupled unsupervised learning with applied factor analysis to the five-factor model (FFM), an empirically-researched taxonomy for personality traits used to describe the human personality and psyche. I programmed these statistical algorithms on an Open Psychometrics dataset with 1,015,342 answered questionnaires based on descriptors of common language and not on neuropsychological experiments for accessibility. Used K-means clustering and feature scaling (min-max normalization) in Python to group certain trait variances into 1 out of 10 possible personality types.

The 10 Personality Types

References

Goldberg, L. R. (1992). The Development of Markers for the Big-Five Factor Structure. Psychological Assessment, 4(1), 26–42. https://doi.org/10.1037/1040-3590.4.1.26