Can you predict Big Five personalities without online data?

I have previously documented my methodologies, connecting human activity data of large groups of people to a series of psycholinguistic based personality traits. 

Recently, we began experimentation with personality traits that are used to calculate a theoretical Big Five personality type Using Spearman’s and Pearson’s correlations

The results are quite different using the different correlation methodologies. It appears Spearman is a better predictor of outcome. My goal is to better predict who will buy things without any online click-based or cookie base data. The same goes for sessions or IP addresses.

Below are some of the early-stage results. Below is a summarization of people throughout the United States who have bought these types of automobiles using Pearson’s correlation. This is not predictive in nature but looking at historical auto data sales connected to human activities of buyers. The calculations do not connect to people but two groups of people, removing personal information in the process.

My next set of reports will examine more detailed data showing some last degrees of differences between cat and dog owners and how it connects to personality traits.

You may also like