About four years ago I first began advocating for the use of personality tests (like Myers Briggs) in intelligence shops both to help identify cognitive biases among analysts and assist in audience analysis to make products more likely to resonate with consumers.
For reasons that are hidden from me, I recently began thinking about this again and asked a number of people working in the intelligence field to indulge me and take an online test to determine their Myers Briggs personality types. I received responses from a total of 23 persons* and here are the results:
First the breakdown into the 16 different personality types.
The two columns of percentages compare the percentages of the whole among respondents (titled ‘% of total) and estimates of the distribution of personality types across the entire population.
Note how skewed the INTJ and ENTJ types are. In my sample, these two types comprise fully 40% of all respondents yet in a random sample of people should make up less than 5% of respondents. Why so overrepresented? Well, let’s look at some of the commonalities and begin with the NT pairing. Using the Keirsey temperament sorter, that pairing is described as ‘rational‘.
As described in the Wikipedia:
As the knowledge-seeking temperament, Rationals trust reason implicitly. They rely on objective observations and factual analysis in any given situation. They seek a logical argument as a basis for action. As strategists, Rationals strive to gain as much information as possible, applying what they learn to develop long-term plans and the steps for achieving them. They are characterized by a tough-minded personal style, tending to pursue either power or understanding.
That sounds pretty much like what you’d want from an analyst, no? But, that personality comes as a bit of a double edged sword if you’re organization is the kind that likes obedience, conformity and/or respect for positional authority.
They are often strong-willed, ambitious, intelligent, and self-determined. Subjective thoughts and emotion have no place in the decision-making process of a Rational. Driven to excel, they work hard to achieve their goals, and they do well where they can take control or work independently on a task.
So, now let’s look at the distribution of the individual characteristics.
Here, the interesting thing is that only in the thinking/feeling dichotomy do the respondents look like the general population. Other than that there are some pretty significant differences.
The extrovert/introvert difference isn’t too surprising to me as I’ve always thought analysts are more introverted than the general population and these respondents seem to confirm my belief.
Sensing/intuition is seriously out of whack with the general public. Its described as:
Sensing (S) Paying attention to physical reality, what I see, hear, touch, taste, and smell. I’m concerned with what is actual, present, current, and real. I notice facts and I remember details that are important to me. I like to see the practical use of things and learn best when I see how to use what I’m learning. Experience speaks to me louder than words.
Intuition (N) Paying the most attention to impressions or the meaning and patterns of the information I get. I would rather learn by thinking a problem through than by hands-on experience. I’m interested in new things and what might be possible, so that I think more about the future than the past. I like to work with symbols or abstract theories, even if I don’t know how I will use them. I remember events more as an impression of what it was like than as actual facts or details of what happened.
I have to admit I would normally think that analysts would conform much more to the general norm here. My hypothesis is that since a significant number of respondents were associated with law enforcement the emphasis on facts and concepts like ‘beyond a reasonable doubt’, ‘reasonable suspicion’, etc. lead to an increased amount of sensing individuals.
The other thing to note is the almost total lack of the ‘perceiving’ trait (the P/J characteristic which is the last of the four letter code). Whereas, in a ‘normal’ population you’d expect roughly 45% of people to have that characteristic only 4% (one respondent) has it.**
The P/J dichotomy is a bit confusing but here’s what they say:
This pair describes whether you extravert (act in the outer world) when you are making decisions or when you are taking in information.
Some general statements that apply to ‘J’ personalities:
- I like to have things decided.
- I appear to be task oriented.
- I like to make lists of things to do.
- I like to get my work done before playing.
- I plan work to avoid rushing just before a deadline.
- Sometimes I focus so much on the goal that I miss new information.
Some general statements that apply to ‘P’ personalities:
- I like to stay open to respond to whatever happens.
- I appear to be loose and casual. I like to keep plans to a minimum.
- I like to approach work as play or mix work and play.
- I work in bursts of energy.
- I am stimulated by an approaching deadline.
- Sometimes I stay open to new information so long I miss making decisions when they are needed.
I find the complete lack of ‘P’ types (other than me) a bit disturbing since I can’t imagine a complete deficit of a particular personality type being good.***
The next step (which, in all honesty, I won’t do because I just don’t have the logistical mojo to make it happen) would be to test supervisors and (more importantly) consumers of intelligence and see their personality types. Then, you could do some really interesting comparisons and perhaps draw some interesting strategies for intelligence production that has a better chance of resonating with the audience.
As of right now, this might have some implications for partnering on projects and how to best introduce new ideas and projects between and within analyst circles. That’s where I’ll try to focus my efforts on next. Stay tuned.
*This isn’t a representative sample but I think you’ll agree the results are interesting nonetheless.
**That one respondent? Well…my mom always did say I was unique.
***I’ve gotten three or four additional responses since writing this and the answers really conform to what I wrote here. I remain the only ‘P’ type out of the bunch which is looking even less like a fluke.