Can personality change? The evidence from long-term studies

Yes, personality changes, in a specific way worth understanding. On average people become more emotionally stable, more conscientious and more agreeable as they age, while their position relative to people their own age changes much less. Someone among the most organized in their group at 25 is likely to be among the most organized at 50, even though the whole group has become a bit more organized in between.

Those are two separate questions, and studies measure them differently. Mean-level change asks whether a group shifts with age. Rank-order stability asks whether the people above average stay above average.

Mean-level change: people mature, mostly between 20 and 40

The most cited synthesis pooled 92 longitudinal samples. Conscientiousness, emotional stability and social dominance (the assertive side of Extraversion) rise on average, especially between 20 and 40. Openness and social vitality rise in adolescence and fall in old age; Agreeableness changes mainly in old age (Roberts, Walton and Viechtbauer, 2006).

A newer meta-analysis of studies published after 2005, with more than 240,000 people for mean-level change, confirms the direction and strengthens it for one trait: from adolescence to old age, emotional stability rises by almost one and a half standard deviations, roughly steadily. Conscientiousness rises by about one standard deviation and drifts back toward adolescent levels in old age, Agreeableness rises by about half a standard deviation, and Openness falls by about half a standard deviation during adulthood (Bleidorn and colleagues, 2022).

This pattern has a name, the maturity principle: on average, the years move people toward the traits that make work, relationships and stress easier to handle. It is an average. In a study that followed 1,795 people from 16 to 66, depending on the trait, between 20% and 60% of participants changed reliably: a large share, but not everyone (Damian and colleagues, 2019).

Your position relative to others: stable, not set in stone

The key figure here is the correlation between two measurements of the same people years apart. A review of 152 studies found it rises with age: .31 in childhood, .54 in the college years, .64 around 30 (Roberts and DelVecchio, 2000). The 2022 meta-analysis corrects the later part of the story: after about 25 it found little evidence that stability keeps rising (Bleidorn and colleagues, 2022). The idea that personality keeps hardening until 50 or 60 does not hold up in the newer data.

Over very long spans stability drops. In the 16-to-66 study, the correlation between the two measurements 50 years apart was .31 after correcting for measurement error (Damian and colleagues, 2019). There is a recognizable link between the teenager and the adult, with plenty of room for individual change.

You may have seen the headline that at 77 you are "a completely different person" from who you were at 14. It comes from a Scottish study that compared teacher ratings at 14 with ratings collected at 77 (Harris and colleagues, 2016). The authors list the limits themselves: the early measures were single questions answered by teachers, the later ones were collected differently decades on, and 174 people took part at 77. It is weak evidence, not proof that personality resets.

Therapy changes traits

The most surprising result comes from interventions. A review of 207 studies, mostly of psychotherapy, found measurable trait change in about six months, with an average effect of .37 standard deviations (Roberts and colleagues, 2017). Emotional stability moved most (.57 standard deviations), then Extraversion (.23). The type of therapy mattered little, and the changes held a year later.

Two caveats. Nearly all participants were in treatment, and traits were self-reported, so part of the change may reflect how people see themselves after therapy. It is still evidence that traits are not fixed.

Is wanting to change enough?

Studies of people trying to change on their own give an honest answer: wanting is not enough, doing helps a little. In a study of 377 students over 15 weeks, each received small weekly challenges tied to the trait they wanted to change. Accepting challenges predicted nothing; completing them did. Completing two Extraversion challenges a week went with almost .2 standard deviations more Extraversion, while accepting challenges without completing them went with relative declines (Hudson and colleagues, 2019). The authors note that the data are correlational and self-reported.

Change is slow, it mostly runs in one direction for the average person, and what counts is what you do rather than what you intend. A Big Five test shows where you are now. Taking it again in a few years is the simplest way to see whether you have moved: the test below takes about 10 minutes.