Does personality change over your lifetime?

Personality changes across adulthood, systematically and in a predictable direction. This is one of the better-established findings in the field, and it contradicts both of the popular positions — that personality is fixed by early adulthood, and that it can be transformed by sufficient motivation.

The maturity principle

Large longitudinal studies and meta-analyses converge on a consistent pattern from roughly the twenties into middle age and beyond:

• Conscientiousness rises, most steeply in the twenties and thirties.
• Agreeableness rises, more gradually, continuing into later life.
• Neuroticism falls, particularly through the twenties and thirties.
• Extraversion shows mixed results — the social dominance aspect tends to rise while the social vitality aspect declines.
• Openness rises modestly through young adulthood, then declines slowly in later life.

Roberts, Walton and Viechtbauer's meta-analysis of 92 longitudinal samples established this pattern, and it has held up in subsequent work. The direction is towards what researchers call maturity: more responsible, more emotionally stable, more considerate. Most people, on average, get easier to live with.

The pattern appears across cultures, which is one of the arguments that it is developmental rather than purely a product of any particular society's expectations.

The two kinds of stability

Personality can be stable in two different senses and the difference explains a lot of apparent contradiction in the literature.

Rank-order stability asks whether people keep their relative position. If you are more conscientious than most of your peers at 30, are you still more conscientious than most of them at 50? This is high — correlations over decade-long intervals typically run in the 0.5 to 0.7 range, and rank-order stability increases with age, plateauing somewhere after 50.

Mean-level change asks whether the whole population shifts. This is where the maturity principle lives. Everyone can become more conscientious while keeping their relative ordering intact.

Both are true at once, which is why "personality is stable" and "personality changes" are both correct claims about different quantities.

What moves it

Age itself, through the maturity pattern above.

Life roles. Entering committed employment, a serious relationship, or parenthood is associated with increases in conscientiousness and emotional stability. The social investment theory account is that role demands pull behaviour, and sustained behaviour change eventually registers as trait change.

Therapy. Roberts and colleagues' 2017 meta-analysis of clinical trials found meaningful trait change, particularly reductions in neuroticism, with effects appearing within weeks and largely holding at follow-up. This is one of the more striking findings in the area: traits move under intervention more than the trait framing implies.

Deliberate effort, modestly. Volitional change studies show that people who want to change a trait can shift it, provided the effort is directed at specific concrete behaviours rather than at the trait as an abstraction. Wanting to be more extraverted does little; committing to specific social behaviours does more.

The pace

Effect sizes for naturalistic change over a decade are moderate. Therapy effects arrive faster but are largest for neuroticism and smaller elsewhere. Nothing in the literature supports transformation over weeks.

This has a practical implication for anyone retaking a test. Measurement error on most instruments is wide enough that a few points of movement between administrations is noise. Real trait change shows up over years, as movement well outside the error band — which is also why a single retest is a poor way to detect it, and why comparing scores taken a month apart mostly tells you about the test rather than about you.

The reading that follows

A trait score is a description of where you currently sit, not a fixed property. It is stable enough to be worth measuring and mutable enough that measuring it again in five years is a different question, not a repeat of the same one.