Looking for Humanity in Data: Understanding Human Psychology from Receptiviti

  by    0   0

Understand people as people. This is not so hard a concept, except business has never traditionally operated this way. They see employees as workers. They look at consumers as potential buyers for their products. Rarely was the value to the consumer, or employee considered… until today.

What I’ve witnessed recently within the advertising and even the HR space, is an uncanny desire to understand more about people’s psychology: their interests, motivations, personality – as it relates to the business… or NOT. The latter, while not entirely new, brings more focus to how business can begin to truly understand people, and use this knowledge to drive value, and eventual benefit to the business.

Welcome Jonathan Kreindler, CEO and Co-Founder of Receptiviti.  Receptiviti is a language-based personality analytics technology company. Receptiviti uncovers personality, emotion and tone from unstructured language like email, social, text, surveys and just about anywhere there’s language. The company’s technology has been used by some of the largest tech companies in the world including Facebook and Intel.

In this episode,  we discuss:

  • James Pennebaker’s (Co-founder and Chief Scientist at Receptiviti) work in the study of written language and as a way to understand people, their emotions, prediction of behaviours.
  • The two classes of words and how this impacts understanding of human psychology.
  • The fallacy of Meyers-Briggs (MBTI Personality Testing) to truly understand personality.
  • How can Receptiviti be applied to the organization?
  • What is the vision for Receptiviti in the next 5-10 years?

You can listen to the podcast here or catch the episode here on Libsyn.

James Pennebaker and the Secret Life of Pronouns

I had a chance to watch James Pennebaker’s, Co-Founder of Receptiviti, TedX talk about the Secret Life of pronouns.

Where it all began: Jamie is a social psychologist and worked with individuals going through life traumas. He is a practitioner and proponent of writing therapy, which is the process of expressing thoughts by putting pen to paper. Jamie has employed this therapy with many patients over many years.

With this large corpus of language data tied to the emotional states of its authors, Jamie wanted to understand word use, the patterns of use and relationships between words/phrases and eventually how this related to how people communicated, and informed various psychological states.

The Two Classes of Words

It’s NOT what you say but how you say it.

Jonathan describes the difference between Content vs Function words and their importance in developing Receptiviti.

Categories of subject, the CONTENT are what people are talking about. Natural language Processing (NLP) and sentiment analysis will typically analyze these themes. However, FUNCTION words, ie the pronouns, preposition, and articles  are the “glue words”. These are the part of language that convey the message and they can uncover things about your personality to a certain degree. As per Jamie:

“Pronouns, articles, prepositions, and a handful of other small, stealthy words reveal parts of your personality, thinking style, emotional state, and connections with others. These words, typically called function words, account for less than one-tenth of 1 percent of your vocabulary but make up almost 60 percent of the words you use.”

How each of us uses words to describe e.g. an object in front of us ie it’s color, shape ie in particular those words that  string together the content will differ from one person to the next.

Case Example: 

I am calling a friend on the telephone. I state, “I will pick you up in front of the station at 5:00 p.m.

This one sentence is deconstructed in the following way:

  • This sentence implies we both know the exact location, which also reveals a pre-existing relationship.
  • The statement underlies that both parties know each other well.

Depending on how people use function words, Receptiviti can detect the kind of relationship two people have.  There are patterns that emerge between people who get along vs. those who don’t. Purely based on language samples, analyzing these function words, irrespective of tone, can provide strong insight into the state of the relationship and where it’s headed.

I questioned Jonathan on the prevalence males vs. females usage of personal pronouns (I, me, my), as well as patterns of phrases being able to detect depressive states.

Jonathan cautioned that the technology provides directional results. They are not absolute. There are arguments that fall on both sides of the coin but overtime with increased data perhaps we will begin to see stronger patterns emerging that draw more conclusive results.

Analyzing Group Dynamics

By understanding people and personalities and the quality of these relationships, Receptiviti can help determine what groups will work well together, as well as identify those within the group who are more likely to exert authority.

I asked Jonathan about MBTI (Myers-Briggs Personality Traits) as a common practice introduced to organizations as part of the team building process: both to understand individual propensities and how to improve chemistry. Jonathan noted that MBTI is a measurement of personality based on self-assessment. Receptiviti, on the contrary, measures personality passively without the individual input. As Jonathan notes:

Ask people a series of questions about themselves, their preferences and what they’re most likely to do and based on their answers. They self select themselves to a certain personality type. I am not disputing the validity of those approaches, however there is always a question of bias.

Language is a social construct. The way you speak, the way you talk, the way you engage with other people — to us, this IS your personality. That’s what people judge you by and a critical part of whether people like you and whether they get along with you.

For an organization to find the those who are the right fit, it matters more what everyone else thinks about those individuals vs. what they think about themselves.

Here is an example of how Receptiviti analyzes text and how it derives the data. In this post about Kanye West: Receptiviti analyzed his VMA Awards acceptance speech. Here is the outcome:


The Big Five Framework

Another way of understanding personality is by looking at the Big Five, a framework that determines the degree to which people are

  1. open
  2. conscientious
  3. extroverted
  4. agreeable
  5. neurotic

Receptiviti includes the variables of the Big Five based on exhibited behaviours, and through certain pathologies to medical problems. It is a bottoms-up approach that allows the team to correlate and understand the patterns of language that are unique to people who have a certain pathologies. The Big Five does not have the ability to do this on its own.

Currently the technology relies on the written word. Through email, customer service scripts, blogs, letters etc across different formats, Receptiviti can be used to gain a strong understanding of individuals and groups.

Receptiviti in the coming decade

Jonathan notes that Receptiviti is a learning platform.

What we want to do is to take a science that had not been accessible, had not been usable and making it more usable and get more utility. We want to learn from one area and see if we can apply it to other areas eg. HR and people analytics.

Some of these findings may be applied to other domains that we have not yet anticipated.

For the Betterment of Society

While I see opportunities to leverage the technology to help prevent crimes by understanding patterns in criminals’ psychologies, Jonathan also noted that there is a tremendous amount of opportunity to identify people who exhibit issues with post traumatic stress and being able to analyze enrolment forms to identify the likelihood those applicants will suffer from PTS down the road.

It will be definitely worth watching Receptiviti progress.

Leave a Reply

Your email address will not be published.

This site uses Akismet to reduce spam. Learn how your comment data is processed.