Psychometrics are a useful set of tools, but relying on surveys and bad data, is a recipe for disaster.

I tried to gain an idea of the number of the more conspicuous aspects of the character by counting in an appropriate dictionary the words used to express them... I examined many pages of its index here and there as samples of the whole, and estimated that it contained fully one thousand words expressive of character, each of which has a separate shade of meaning, while each shares a large part of its meaning with some of the rest. — Francis Galton, Measurement of Character, 1884

Sir Galton's words in 1884 begin the long journey of computational linguistics and the lexical hypothesis. These big topics are what led us to a disappointing story about Trump, Facebook, Cambridge Analytica bad math and weak analysis.

The words we use help us understand the human mind. Even surveys can give interpretation to our personality traits, when done correctly. Most research on this subject was university-based and limited in scope for years. No one was surveying millions of Facebook users and applying big machines to understand and cluster personality traits at scale. That changed recently with the advent of access to data, speed of the connection and capacity of the machines.

Surveys are not who we are but how we react to a variety of events at a certain point in time. The results of a survey, once fed into psychometric software will often produce ever greater, amplified error. Is it a good marketing strategy to rely on surveys you send to customers after a successful purchase, or a bad customer service experience? Emotions will throw the results you wish for, while the data you have is ....junk.

Setting aside the bad ethical issues, I have studied what Cambridge Analytica attempted technically and find it problematic. Here is why: using survey data is very difficult. Everyone reading this has completed a survey and depending on the questions and the order, your response can be result in very different results. I have studied social data, survey results, blogs and articles and find the inconsistencies in results unreliable.

When psychometrics is performed ethically, with legitimate data sources, you don't need to supplement it with stories of thugs and Ukrainian prostitutes (See Channel 4's investigative report on Cambridge Analytica).

Psychometrics, when used ethically, is a powerful way to create value and balance. When manipulated by fools, it is a disaster.