You know yourself less well than you think, which is why self-assessments are unreliable

The illusion of self-awareness

Self-awareness is one of the most overrated abilities we have, in the literal sense of the word. In organizational psychologist Tasha Eurich's research program, an overwhelming majority of participants rated themselves as self-aware, while only around 10 to 15 percent met the researchers' criteria for actual self-awareness. The pattern recurs in study after study: we trust our inner image of ourselves, but that inner image is never examined by anyone but ourselves. It is like grading your own exam, with an answer key written by the same person who wrote the answers.

The phenomenon is not limited to particular individuals or particular settings. Psychologist David Dunning and his colleagues have shown that flawed self-assessments follow us through most domains of life: at work, in education, in health decisions. And those who rate themselves most generously are often the ones with the least justification for it.

The adaptive unconscious, or why introspection is not enough

Perhaps the most influential explanation of why self-assessments fall short comes from social psychologist Timothy Wilson. In his book Strangers to Ourselves (2002) he describes what he calls the adaptive unconscious: an extensive system of mental processes that interprets the world, evaluates situations and initiates behavior, entirely without passing through consciousness. The adaptive unconscious is not a Freudian basement of repressed drives, but a highly efficient autopilot that allows us to function at all in a complex world.

The problem is that the autopilot keeps no logbooks. When we introspect, looking inward to understand why we do what we do, we do not gain access to the processes that actually drove the behavior. We gain access to a story about them. Wilson's conclusion is as simple as it is uncomfortable: anyone who wants to know themselves cannot only dig inward, because most of the material cannot be dug up. Instead, we need to study ourselves much as we study other people: by observing what we actually do, and by learning how others perceive us.

Telling more than we can know

As early as 1977, Richard Nisbett and Timothy Wilson showed in a now classic article that people rarely have direct insight into the causes of their own choices. In a series of experiments, participants made decisions that were in fact influenced by factors they were completely unaware of, such as where in a display an item happened to be placed. When asked to explain their choices, they confidently gave entirely different reasons: quality, feel, preference. No one said "I chose it because it was furthest to the right", even though that was exactly what had happened.

This is called confabulation: the brain constructs a plausible explanation after the fact and presents it to us as if it were a memory of the actual cause. We experience the explanation as true, which is exactly what makes it so treacherous. In that sense, a self-assessment is not a report from the inside. It is a theory about ourselves, formulated by a theorist with limited access to data and a strong vested interest in the conclusion.

When self-image meets the outside view

So what happens when people's self-assessments are compared with how others rate the same behavior? The answer: they diverge, often substantially. In a meta-analysis by Heidemeier and Moser (2009), the relationship between self-rated and supervisor-rated job performance was weak, a correlation of around 0.22, which in practice means that the self-image and the outside view are largely two different pictures.

At the same time, the picture is more nuanced than "others always know best". Psychologist Simine Vazire's SOKA model (Self-Other Knowledge Asymmetry) shows that we ourselves have the best insight into our inner states (worry, intentions, feelings), while the people around us are often more accurate judges of our observable behavior and of traits loaded with prestige, where our own judgment is easily colored by wishful thinking. This is an important clarification: the question is not whether self-assessments are worthless, but what they actually measure. They measure the self-image. If you want to know how behavior lands with others, you have to ask the people it lands on.

The self-assessment is not worthless, it is a point of comparison

Here lies a point that is often lost in the debate. The fact that self-assessments are unreliable as descriptions of reality does not make them uninteresting. On the contrary: the gap between how I see myself and how others see me is in itself one of the most useful development insights there is. Research on leadership assessments suggests that it is precisely the discrepancy that matters: leaders who consistently rate themselves higher than the people around them do tend to have a harder time developing, because they see no need to.

In other words, the self-image is not the answer key, but it is a necessary half of a comparison. The other half has to come from the outside.

The road to self-awareness leads outward

If introspection is not enough, what is? Wilson's own answer, and that of the research as a whole, points to two sources. The first is our own behavior: observing over time what we actually do, rather than what we think we would do. The second, and most powerful, is other people's perspectives. External feedback is what actually creates awareness, because it contains information that by definition is not available from within: how behavior is perceived, interpreted, and how it affects others.

That is why structured feedback, where several people who see us in everyday life describe our behavior, consistently proves to give more accurate pictures than self-reporting alone. The mirror needs to be held up by someone else. In practice, this means something concrete for everyone who works with developing leaders and groups: tools and processes built solely on self-assessment measure the self-image, not the behavior. If you want to work with how a person is actually perceived, and with the gap between self-image and the outside view, the outside perspective has to be built into the method itself.

That is also the core idea behind IDI, the Interpersonal Dynamics Inventory. An IDI profile is based on how colleagues and coworkers actually perceive a person's behavior, not solely on the person's own self-image. The difference between the two thus becomes not a source of error but the very starting point for development.