VZ editorial frame
Read this piece through one operating lens: AI does not automate first, it amplifies first. If the underlying decision architecture is clear, AI scales clarity. If it is noisy, AI scales noise and cost.
VZ Lens
In VZ framing, the point is not novelty but decision quality under uncertainty. According to William Fleeson’s research, personality is not a fixed type but a distribution of behaviors. This overturns the traditional concept of persona construction and replaces it with a flexible model. The real leverage is in explicit sequencing, ownership, and measurable iteration.
What We Learned from William Fleeson — and Why Does This Undermine Traditional Persona Logic?
TL;DR
When someone says their personality is “introverted” or “anxious,” they’re actually describing an average. But behind that average lies a wide range of variation. William Fleeson’s 2001 research showed that the same person exhibits enormous behavioral differences from day to day and situation to situation—and this should be viewed not as a personality disorder, but as normal functioning. This finding overturns the logic of classic persona-building—and shows what we should do instead.
A Morning in Vienna
I’m sitting at a table in a Viennese coffeehouse, in the quiet of the morning. The aroma of coffee mingles with the sensory memory of old wood and a vacuumed carpet. Before me lie an empty notebook and a pen. My fellow researchers are still quietly lost in their own thoughts; only the clinking of cups breaks the peace. I look at the blank page waiting before me and think about how differently we behave on a calm morning like this compared to a crowded conference room or a dinner with friends. The question circling in my head isn’t who I am here, in this silence. It’s where all those other “me’s” are—the ones that appear and disappear day after day.
1. That particular experiment
William Fleeson is a psychologist at Durham University. In 2001, he published a study that became one of the most cited works in personality psychology.
The experiment was simple. Participants were asked to rate themselves several times a day, over the course of weeks: what state are they in right now? How extroverted, open, or conscientious do they feel?
The results were surprising.
They did not find that everyone remained stable on their own personality scale. Instead, they found that every single person moved significantly along the scale. Even strong introverts had days when they were distinctly extroverted. Even those with low neuroticism scores experienced intense anxiety in certain situations.
A personality trait isn’t a fixed point. It’s a distribution: there’s a mean, but there’s also a standard deviation.
2. What does the distribution mean?
Imagine that Katalin’s neuroticism score is 0.45—a midpoint value, indicating moderate emotional sensitivity.
According to the classical view, this means: Katalin is a person with moderate anxiety. If we ask her a question, her answer will fall somewhere in the middle range.
But according to Fleeson’s findings, the picture is different. Katalin is not always moderately anxious. On some days, she’s at 0.2 (very stable, calm, focused). On other days, she’s at 0.8 (severe anxiety, rumination, decision paralysis).
0.45 is the average. It is the standard deviation that determines how she behaves in a specific situation.
[!NOTE] The Fleeson Principle A personality trait is not a single point on a scale, but a distribution (mean + standard deviation). The situation, stress, context, and social space determine where someone stands on this distribution at any given moment.
3. Why is this crucial from a market research perspective?
The classic persona represents a single point. “Katalin has moderate anxiety.” This single number encapsulates every decision-making situation she might face.
But reality is different. Katalin is different when:
- she sits in front of the computer feeling fresh and rested
- she receives a payment reminder via email after an eight-hour workday
- a friend tells her they had a bad experience with the brand
- the product she relies on is discontinued
In these situations, you don’t get a score for the exact same point. You get four different scores across the distribution.
If a market researcher asks, “How does Katalin react to the new pricing?”—it’s not enough to know where the average lies. You need to know which situation we’re in, and in which direction that pushes Katalin on her distribution.
4. The Three Propositions of Whole Trait Theory
Fleeson’s 2004 extended theory—which he named Whole Trait Theory—formulates three important propositions:
Proposition 1: A trait is both descriptive and causal. High neuroticism does not just mean that someone is “generally more anxious”—it also means that they actively create different situations for themselves and perceive the same situation differently.
Proposition 2: A trait is comparable across different people. The distribution-based approach preserves interpersonal comparability. The means of two people are comparable even if both have standard deviations.
Proposition 3: The trait-based and state-based approaches are not contradictory but complementary. A trait tells us where the mean of the distribution lies. A state tells us where the person is right now, in this particular situation.
5. Standard Deviation as Research Data
When a market research team reads the persona question with Fleeson, a very specific question arises: what is the standard deviation for that dimension, and what triggers the extreme values?
This question is not abstract. It is very practical:
| Dimension | Mean | Triggered by low state | Triggered by high state |
|---|---|---|---|
| Neuroticism | 0.45 | Routine, stable circumstances, good sleep | Stress, unexpected change, uncertainty |
| Conscientiousness | 0.65 | Creative, open-ended tasks, relaxation | Deadlines, supervision, responsibility |
| Extraversion | 0.40 | Working alone, exhaustion | Group tasks, recognition, lively environment |
If we know what triggers extreme values, then we can tailor market research scenarios accordingly—not to an idealized consumer state, but to real-life situations.
6. Intraindividual variability — a person’s own variance
One of the most important concepts in Fleeson’s work is intraindividual variability: the internal behavioral variability of an individual.
This is not a pathology. It does not mean that someone is unstable or unpredictable. On the contrary: it is a sign of normal human flexibility.
People are capable of adapting to the situation. A person with high intraindividual variability is more flexible—they can be both introverted and extroverted, depending on what the situation calls for. A person with low variability is more rigid—they are the same everywhere, which is an advantage in some situations and a disadvantage in others.
From a market research perspective, this means: some people react similarly in a wide variety of situations (low variability), while others exhibit strongly different patterns in different contexts (high variability).
If we ignore this dimension—and the classic persona does ignore it—then the persona presents a false picture of human flexibility.
7. What needs to change in persona development?
Fleeson’s theory does not say that we should discard trait-based thinking. It says: let’s supplement it with state dynamics.
In practice, this means three changes:
1. Don’t store a single point; store a range. Don’t label it “moderate anxiety,” but rather “moderate on average, but shifts quickly toward high anxiety under stress.”
2. Model the impact of the situation. What kind of situation pushes the person toward the high end? What pushes them toward the low end? These cannot be guessed—they must be researched.
3. Take the activation logic into account. It’s not just a question of whether someone is high or low in something, but rather what triggers the higher or lower state. This trigger logic is the engine of the dynamic persona.
8. The Trap of Type Thinking
Traditional typologies—MBTI, DISC, Enneagram—are appealing because they provide structure. A type label condenses a great deal of information.
But this is precisely their weakness: they condense, they do not break things down. An INFJ type label says nothing about how that person behaves when they feel their identity is threatened, when they face financial stress, or when everyone else in a group has a different opinion.
The type provides a mean value. It does not account for variance, activation logic, or context sensitivity.
From Fleeson’s perspective, the type isn’t bad—it’s just not enough. It’s a good communication tool in workshops. But it’s not a good predictive model for critical research decisions.
[!WARNING] The MBTI Trap If your team bases personas on MBTI types, don’t throw them out—but add context: in what situations do they behave this way, and in what situations do they behave differently? The type is just a signpost, not a map.
9. What does this look like in practice?
In a dynamic persona system, Katalin’s description doesn’t look like this:
“Katalin is a woman with moderate anxiety and high conscientiousness.”
But rather like this:
“Katalin’s average neuroticism score is 0.45, but she has high sensitivity to uncertainty (IoU: 0.70). Under stable conditions (low threat to control), it hovers around 0.25; under stress (deadline + unexpected change simultaneously), it can rise above 0.75. In this state, she avoids making decisions, delegates to others, and requests more information before acting.”
This is not a character profile. It is a state model. Fleeson’s theory provides the theoretical foundation.
10. Summary
William Fleeson’s research showed that a personality trait is not a single point but a distribution. Every person has a mean and a standard deviation—and the standard deviation is just as important as the mean.
From the perspective of market research and synthetic persona building, this means: we should not model types, but states, and the situations that activate these states.
This shift in perspective is the foundation of what we will explain in detail in the following articles: the Big Five model, the trait-state distinction, stress dynamics, and CAPS behavioral signatures.
This article is the third part of the Synthetic Personas series. Next up: The Big Five — why is it the gold standard in personality research?
Zoltán Varga | vargazoltan.ai — Market research, artificial intelligence, synthetic thinking
Strategic Synthesis
- Translate the thesis into one operating rule your team can apply immediately.
- Monitor one outcome metric and one quality metric in parallel.
- Run a short feedback cycle: measure, refine, and re-prioritize based on evidence.
Next step
If you want your brand to be represented with context quality and citation strength in AI systems, start with a practical baseline and a priority sequence.