What 50 Years of Personality Research Taught Us About Change

In 1980, a research team at the Baltimore Longitudinal Study of Aging started tracking a few hundred adults on a personality battery. Some of those same people are still in the study today, 45 years later. Their files are one of the longest continuous records of human personality ever collected. The data on those files has quietly settled a question popular psychology has been arguing about since the 1970s: do people change?

The answer, after five decades of work by Costa, McCrae, Roberts, Mroczek, and a few hundred collaborators, is more interesting than a clean yes or no. People are remarkably consistent. People also change. Both are true, and the research has been careful enough to say when.

This is a tour of what the long studies actually found.

Finding one: traits are stable, but not frozen

Costa and McCrae's work through the 1980s and 1990s showed something that surprised the field: rank-order stability for Big Five traits is around 0.6 to 0.8 over decades ¹. That is, if you are higher on Conscientiousness than your friend at 30, you are very likely still higher than your friend at 60.

But rank-order stability is not the same as mean-level stability. The same person can keep their place in the population while still drifting in absolute terms. Roberts and colleagues' 2006 meta-analysis of 92 longitudinal studies found reliable mean-level change across the adult lifespan, particularly between ages 20 and 40 ².

The popular version of this finding — "personality is set like plaster by 30" — turns out to be wrong. The plaster sets, but not as early as Costa originally claimed, and not as completely.

Finding two: most people get a little better

The pattern of average change across adulthood is remarkably consistent across cultures and study designs. People tend to get higher on Conscientiousness, higher on Agreeableness, lower on Neuroticism. Extraversion and Openness drift more modestly and in different directions for different subgroups.

Roberts and Mroczek called this the "maturity principle" ³. Across populations, the trait shifts that come with age are the shifts that make people more functional inside societies — more reliable, more cooperative, less reactive. The shift is not dramatic. About half a standard deviation across 30 years. But it is real and it is in the same direction for most people.

The cynical reading is that we get duller with age. The more accurate reading is that most people slowly become better partners, employees, and parents, whether or not they are trying.

Finding three: change is biggest in your twenties

If you graphed personality change across the lifespan, the steepest slope would be in the 20s and early 30s. Roberts and colleagues' meta-analysis found that the largest mean-level shifts in Conscientiousness and emotional stability happen in this window ².

Why this window matters: it overlaps with the period when most people are entering long-term jobs, partnerships, and parenthood for the first time. The social investment theory argues that personality change is partly downstream of the new roles a person takes on. The job demands more reliability; reliability slowly grows. The partnership demands more patience; patience slowly grows.

This may also be why people in their late 20s often feel like a different person than they did at 22. They are. About half a standard deviation different, on average.

Finding four: change can also go the wrong way

The maturity principle is an average. Inside the average, plenty of people move the other direction. Mroczek and Spiro's work found that increases in Neuroticism across midlife were a meaningful predictor of mortality, independent of physical health markers ⁴.

The personality literature is unusually careful about this. Change is possible. Change is not automatically good. Some people grow more reactive, more disagreeable, more disorganized as life squeezes them. The same flexibility that lets traits improve also lets them deteriorate.

This is part of why "personality is destiny" reads as wrong to people who have seen close family members change for the worse. The data backs them up.

Finding five: intentional change works, modestly

The most recent wave of research — Hudson, Roberts, and a small but growing list of intervention studies — asks the question popular psychology assumed it already knew the answer to: can you change your own personality on purpose?

Roughly, yes, but less than self-help books suggest. Hudson's intervention studies show that people who set explicit goals to change a specific trait (say, Conscientiousness) and engage in repeated trait-relevant behavior over 16 weeks shift their scores by a measurable but small amount ⁵. The change is real. The change is not transformational.

The honest framing: personality change is more like fitness than like enlightenment. You can move the dial. The dial does not flip.

Finding six: tests can capture this, but only if you use them right

A single personality test taken once gives a snapshot. The thing the longitudinal studies clarified is that the snapshot is informative, but it is not a sentence. A test result at 25 should be expected to drift by 30 and again by 40. Anyone using a test as a permanent label is using it wrong.

The more useful frame: take the test, get the snapshot, build something with it for the next two to three years, and then take it again. The change between snapshots is often more interesting than either snapshot alone.

What all of this adds up to

Fifty years of careful work converges on a small set of claims that are worth holding lightly and using often.

People are mostly consistent. The you of next year will look a lot like the you of last year on a personality test.

People also change. Across decades, the average person becomes a little more reliable, a little kinder, a little calmer. Across shorter windows, with effort, smaller shifts are possible.

The biggest change happens in your 20s, often without trying. After that, change exists but takes more deliberate work.

And the test result is a starting point, not an ending. The most interesting thing a personality test can do for you is give you a baseline to revise.

Take the Big Five test and set your own baseline →

References