Personality Traits

Does Personality Change? What Longitudinal Data Shows

PsyCove Editorial · 8 min read · Updated 2026-10-09

Yes, but slowly and in specific ways. Personality shows high rank-order stability — people tend to keep their relative position — while also showing reliable mean-level change: on average, people become more conscientious, more agreeable, and more emotionally stable from young adulthood into middle age. Both are true at once, and confusing them is the source of most bad arguments about whether change is possible.

Key takeaways

Personality shows high rank-order stability and real mean-level change at the same time; these are different questions, not competing answers.
Rank-order consistency rises from roughly 0.3 in childhood to around 0.7 later in life, so change gets slower with age without ever stopping.
Mean-level change is largest in young adulthood and amounts to a fraction of a standard deviation per decade — real, but slow.
Interventions move traits most reliably on emotional stability, and they work better by changing daily demands than by changing insight alone.
Track trait change over months and years; weekly re-scoring measures noise, and switching instruments mid-stream makes the comparison meaningless.

Two kinds of stability

Rank-order stability: do people keep their relative position in a group? This is high, and it rises with age — which is why your scores correlate strongly with your own scores a decade later.
Mean-level change: does the average level of a trait in a population shift? This happens reliably, and it is not an artifact of measurement.
Individual change: some people change substantially in ways the averages hide. Averages describe groups; they do not cap individuals.

You can be recognizably the same person and measurably different. Stability and change are not rivals — they are two different questions.

What the maturity principle says

Across many samples, young adulthood into middle age brings increases in conscientiousness and agreeableness, and decreases in neuroticism. This pattern is often called the maturity principle: the traits that make social roles work — reliability, cooperation, emotional steadiness — tend to grow as people take on those roles and stay in them. It is an average, not a guarantee, and the size of the shift is moderate rather than dramatic.

What actually drives change

Sustained roles: a job, a partnership, or a caregiving role that rewards a trait consistently will pull behavior, and behavior repeated becomes trait-relevant.
Life events with lasting structure: not the event itself but the daily demands it installs.
Deliberate practice with feedback: change is fastest where the behavior is specific, the feedback is immediate, and the environment keeps rewarding the new pattern.
Reduced load: chronic stress narrows behavioral range; lowering it restores the capacity to choose differently.

How to work on a trait honestly

The realistic target is not a new personality but a shifted distribution of your behavior: you act differently often enough that others would describe you differently, and that your own default moves at the margin. Targets should be concrete and environmental rather than global and motivational.

Pick the behavior, not the trait: 'I want to be more conscientious' fails; 'I close my day by writing tomorrow's first task' works.
Change the environment before the will: make the desired action the default and the undesired one effortful.
Use commitment devices that survive a bad day — the plan has to work when motivation is absent, because most days are like that.
Measure in months. Re-scoring a trait inventory quarterly is a reasonable cadence; weekly is noise.

How much do personality scores actually move?

Rank-order stability, decade by decade

Rank-order stability asks whether people keep their relative position, and the answer depends heavily on age. Roberts and DelVecchio's quantitative review of longitudinal studies spanning childhood to old age found consistency estimates of roughly 0.3 in childhood, in the 0.5s through the college years, and around 0.7 in the second half of life. Consistency also falls as the interval between measurements lengthens, which is what you would expect if change is cumulative rather than sudden. The practical reading is that a trait score predicts your relative standing years later — not perfectly, but far better than chance.

How large is mean-level change?

Mean-level change asks whether the population average moves, and it does. The largest shifts appear in young adulthood and early middle age, where conscientiousness, agreeableness, and emotional stability all rise on average. The size is a fraction of a standard deviation per decade, which sounds negligible and is not: at that rate a cohort's average profile after twenty years looks measurably different, even though no individual can point to the week it happened. After roughly the age of thirty the same shifts continue, but more slowly.

Why stability gets stronger with age

Three mechanisms are usually offered, and they are not mutually exclusive. The first is accumulation: by middle age you have spent decades in environments you partly selected, and those environments keep reinforcing the same tendencies. The second is identity consolidation — once you and the people around you hold a settled account of who you are, behavior that contradicts it costs more socially. The third is declining exposure to novel role demands, since the twenties install more new roles per year than any decade that follows.

None of this makes change impossible, and the evidence supports keeping the two apart. Longitudinal work finds individual-level change well beyond what group averages imply, including a minority of people who move against the maturity trend. The mechanism worth remembering is that stability is partly manufactured by stable circumstances. Alter the circumstances durably — a new role, a different city, a sustained practice — and the pressure toward consistency relaxes.

Can therapy or training change personality? What the trials show

This is a question the literature has taken seriously only relatively recently, and the answer is cautiously positive. A meta-analysis of 207 intervention studies, clinical and non-clinical, put the average trait change at d = .37 over an average of 24 weeks, with the largest and most consistent movement in emotional stability. The caveat is one of scale: the changes are modest, they vary considerably between studies, and many trials are small and short. It is fair to say the evidence establishes malleability without supporting any promise about magnitude.

Clinical treatment: reductions in neuroticism are among the better-documented effects of effective therapy, and they tend to arrive alongside symptom improvement rather than separately from it.
Non-clinical programs: structured training and coaching studies report smaller but non-zero shifts, concentrated on conscientiousness and emotional stability.
Duration matters: longer and more intensive programs generally produce larger effects than brief ones, which argues against weekend interventions of any kind.
Openness and agreeableness are the least responsive domains in most reviews, which suggests some traits simply offer fewer handles than others.
Follow-up is the weak point: relatively few studies measure whether change persists beyond a year, so durability remains the least settled part of the claim.

How to change a trait on purpose

If the mechanism is repeated behavior under stable conditions, the protocol follows directly: make the behavior specific, make the conditions durable, and measure on a timescale long enough for a real signal to appear. Three steps cover most of what the evidence supports, and none of them require you to feel like a different person first.

Pick a behavior, not a trait

'I want to be more conscientious' has no action attached to it, so there is nothing to repeat and nothing to notice. 'I write tomorrow's first task before I close the laptop' is repeatable, observable, and falsifiable within a day. The principle generalizes: a trait is an aggregate of behavior, so the unit of change is a behavior that recurs often enough to move the aggregate. Choose one that happens daily rather than one that happens when you happen to feel motivated.

Change the conditions before the will

Willpower is a poor plan because it is least available exactly when you need it. Environmental design is more reliable: make the wanted action the default and the unwanted one effortful, put the cue where the behavior has to occur, and remove the friction that currently protects the old pattern. This is also why role changes do more work than resolutions — a job with hard external deadlines does not require you to feel conscientious in order to act conscientiously, day after day, until the behavior starts to look like a trait.

Measure on a timescale that can show a signal

Trait scores are stable enough that weekly re-testing mostly measures noise, and re-testing with a different instrument each time destroys the comparison. A quarterly cadence with the same inventory is a reasonable compromise: frequent enough to detect movement within a year, infrequent enough that you are not chasing sampling error. Pair the score with a behavioral count — how many days the target action actually happened — because that count moves earlier than the trait does and tells you whether the protocol is running at all.

Where the evidence is mixed

Two claims in this area are weaker than their popularity suggests. The first is that specific events — a marriage, a redundancy, a bereavement — reliably change personality; the better evidence says events matter through the daily structure they install, and that an event with no lasting structural consequence produces a shift that decays. The second is that intentional personality change is straightforwardly achievable; the trials show change is real, but the effects are modest and unevenly distributed across people.

Events versus structure: the same event produces different outcomes depending on what daily demands follow it, which is why averaging across event types hides the effect.
Which traits move: emotional stability and conscientiousness have the most support; openness and agreeableness have considerably less, and this is contested.
How long change lasts: follow-up data beyond twelve months is sparse, so confident claims about permanence are not well grounded.
Self-report inflation: most intervention studies rely on self-report, and people who invest in changing a trait are motivated to report that it changed.

Change is real, slow, and uneven — and it is produced by the unglamorous repetition of one specific action under conditions that keep asking for it.

Related types

🔪 The Chisel

Related guides

Frequently asked questions

Can therapy change personality?

It can, particularly for neuroticism, where reductions in negative emotionality are among the better-documented effects of effective treatment. The change is usually gradual and tracks symptom improvement rather than arriving separately from it.

Do big life events change you permanently?

Rarely by themselves. Events matter mainly through the daily structure they install afterwards. An event with no lasting change in daily demands tends to produce a temporary shift that decays.

Am I too old to change?

No. Stability rises with age, which means change gets slower rather than impossible. The practical implication is to allow more time and to rely more heavily on environment than on motivation.

How long does it take to change a personality trait?

Months at minimum for the underlying behavior, and typically one to three years before a self-report trait score moves enough to separate from noise. Brief interventions of a few weeks are generally too short; reviews find longer programs produce larger effects.

Which personality traits change the most?

Emotional stability and conscientiousness have the strongest evidence for change, both through maturation and through intervention. Openness and agreeableness show the least movement in most longitudinal and intervention reviews, though this is contested rather than settled.

References

This guide summarizes published research. It is not a diagnosis, not therapy, and not a hiring, admission, or screening instrument. PsyCove Editorial is an in-house editorial team, not a clinical practice; if distress is persistent, contact a qualified professional or a local crisis line. PsyCove is not affiliated with or endorsed by the MBTI® trademark.