We underestimate how much personality changes with age, study finds

By Tarun Sai LomteReviewed by Susha Cheriyedath, M.Sc.Jul 29 2026

Comparisons spanning ages 15 to 90 challenge the enduring belief that adult personality remains largely fixed.

Study: Personality traits are less stable than people think. Image Credit: Ribkhan / Shutterstock

In a recent study published online as an 'article in press' in the journal Communications Psychology, researchers compared changes in personality traits with changes in other aspects of life.

Personality traits were historically regarded as largely stable and immutable once people reached adulthood. Previous research suggests that personality traits develop throughout the lifespan and provides promising initial evidence that targeted interventions can modify them. The notion that personality is highly stable persists among the public and in academia. The extent to which personality traits change compared with other aspects of life remains unclear.

About the study

In the present study, researchers compared lifespan shifts in personality traits with those in other aspects of life. They conducted two studies; the first examined whether laypeople believed that personality traits change less over time than variables from different domains of life. The second study empirically assessed whether population-level mean levels of personality traits change less across the lifespan than those of other variables.

For the first study, the final sample comprised 887 laypeople and was designed to be representative of the United States (US) population by age, sex, and race. Laypeople were those without a doctoral degree in psychology. Participants rated how much they believed the population mean level of a given variable would change, on average, from ages 15 to 90 on an 11-point sliding scale, with a separate “Not at all” option coded as zero. Specifically, there were 25 variables, including the Big Five personality traits, health indicators, cognitive ability, income, engagement, leisure, and social variables.

The team compared participants’ ratings of all individual variables and aggregated ratings by domain, for example, health and personality traits. In the second study, they leveraged secondary data from eight longitudinal studies conducted in the US, Germany, the Netherlands, and Australia. These studies were representative panel studies that each had at least three waves of Big Five data and encompassed 166,971 participants, all of whom had at least one wave of Big Five data.

The researchers examined the Big Five alongside more than 30 other individual-difference variables across health, finance/career, relationships, self and identity, engagement, leisure, and social functioning domains. The principal standardized comparisons involved 27 non-personality variables. The team modeled population-level lifespan mean trajectories of these variables using hierarchical age-period-cohort (HAPC) cross-classified random-effects models (CCREMs). Finally, they derived per-variable meta-analytic lifespan trajectories and compared perceived and empirical lifespan changes.

Findings

Laypeople’s ratings of perceived lifespan change in the Big Five personality traits ranged between 4.58 and 5.87. Laypeople rated neuroticism and extraversion as the most stable among the Big Five traits. Conscientiousness received the highest relative change rating. Locus of control and self-esteem had average change ratings. The self and identity domain was rated as having average changes. By contrast, the Big Five domain was rated as having the least change.

All Big Five traits were rated as changing significantly less than cognitive ability, self-esteem, locus of control, loneliness, life satisfaction, physical health problems, health status, health limitations, income, sleep, anxiety symptoms, and the frequency of leisure, exercise, and socializing. Participants rated the Big Five traits as changing by 0.43 to 0.93 standard deviations less relative to their ratings of the other 20 variables.

Laypeople rated the Big Five domain as significantly more stable than other domains. Net change compared estimated values at ages 15 and 90, while cumulative change summed the magnitude of all estimated year-to-year shifts regardless of direction. In the second study, among the Big Five traits, the estimated mean levels of openness, neuroticism, and extraversion decreased overall from ages 15 to 90, whereas conscientiousness and agreeableness increased overall. Among 27 other variables, positive affect, leisure frequency, exercise frequency, and health status showed the largest net reductions.

In contrast, health limitations, body mass index, physical health problems, and satisfaction with finances showed the largest net increases. The Big Five traits showed larger or similar magnitudes of net change than many variables. Average yearly cumulative changes ranged between 0.017 and 0.026 standard deviations for the Big Five traits, corresponding to total cumulative changes of approximately 1.29 to 1.91 standard deviations from ages 15 to 90. Across variables, the largest cumulative lifespan changes were observed for socializing frequency, income, health limitations, sleep, and conscientiousness.

Meanwhile, the smallest cumulative changes were in the importance of religion and satisfaction with finances and family. Collectively, lifespan cumulative changes in personality traits were comparable to, or even larger than, multiple common variables.

Finally, laypeople’s perceptions of lifespan change broadly aligned with meta-analytic estimates, though they underestimated changes in personality traits relative to other variables.

Conclusions

In sum, population means levels of personality traits are less stable than commonly believed. Laypeople underestimated the extent of personality trait development compared to other life aspects; longitudinal estimates indicate that these traits have mean-level changes comparable in magnitude to those of other aspects of life and continue to develop throughout the lifespan.

The modeled population-average trajectories do not show that every individual changes in the same way, represent the same people followed continuously from ages 15 to 90, or address whether people retain their relative personality rankings over time. All data came from Western democratic countries, which may limit the generalizability of the findings. 

These results highlight a persistent gap between empirical evidence and public perceptions of personality traits, and effective communication about their malleability may help close it.