Image: Pedro Ribeiro Simões, CC BY 2.0, via Wikimedia CommonsIs personality fixed after 30
"In most of us, by the age of thirty, the character has set like plaster, and will never soften again." William James wrote that in The Principles of Psychology in 1890, and for most of the twentieth century the main theorists of personality agreed with him. The strongest statement came from Paul Costa and Robert McCrae, the authors of the most used Big Five questionnaire, who argued in the 1990s that traits are essentially fixed after about 30. The data that have accumulated since then say something more interesting, and the answer depends on which of two questions you ask.
The five traits
Most of the research measures the Big Five: extraversion, agreeableness, conscientiousness, neuroticism (or its opposite, emotional stability), and openness to experience. They were found by asking what the thousands of personality words in the dictionary have in common when people rate themselves and others, and they appear in every language and culture that has been checked, with some argument about the fifth. Each is a continuous dimension, not a type, and most people are near the middle of most of them. Unit 11 of the masterclass gives the method and the evidence.
Question one: do people keep their rank
The first kind of stability is rank-order stability: if you were more extraverted than most of your classmates at 20, are you more extraverted than most of them at 50? Brent Roberts and Wendy DelVecchio answered it in 2000 with a meta-analysis of 152 longitudinal studies, published in Psychological Bulletin and at the journal. The correlation between a person's trait score and their score about seven years later rises steadily across the lifespan: about 0.3 in childhood, 0.5 in the student years, 0.6 in the thirties, and about 0.75 between 50 and 70. It never reaches 1. People keep their rank more and more as they age, and never completely.
So James was right that personality becomes more stable with age, and wrong that it sets at 30. The curve keeps rising into the sixties, and even a correlation of 0.75 over seven years leaves room for substantial change in individuals, especially over longer intervals. Across 40 years the correlation is lower, and some people in every cohort change a great deal.
Question two: does everyone move together
The second kind of stability is mean-level change: does the average person become more or less of something as they age? Roberts, Kate Walton, and Wolfgang Viechtbauer answered it in 2006 with 92 longitudinal studies covering ages 10 to 101, in a paper at the same journal. The average person becomes more conscientious, more agreeable, and more emotionally stable from adolescence through middle age, with the largest changes between 20 and 40, which is the opposite of the plaster hypothesis. The social dominance side of extraversion rises in young adulthood; the social vitality side, and openness, decline in old age. The changes are not small: the gain in conscientiousness from 20 to 60 is about one standard deviation, which means that the average 60-year-old is more conscientious than about 84% of 20-year-olds.
Roberts calls this the maturity principle. The direction of change is the one that makes people better at the roles adult life hands them, as workers, partners, and parents, and the studies that have looked suggest that the roles cause part of the change: people who take a steady job early become more conscientious sooner, and people who enter a lasting relationship become more emotionally stable. The pattern appears in every country studied, which suggests something general about adult life rather than about one culture.
Both questions can be true at once. Everyone moves up on conscientiousness, and the people who were above average at 20 are mostly still above average at 50. The whole cohort climbs the hill in rank order, loosely.
Why it matters what the score is
Traits are worth measuring because they predict outcomes that matter, with correlations in the same range as the variables that get more attention. Roberts and colleagues reviewed the evidence in 2007 in a paper free at PubMed Central: conscientiousness predicts length of life about as well as socioeconomic status does and better than IQ, neuroticism predicts divorce and depression, and the traits together predict occupational success and mortality with effect sizes comparable to those of intelligence and family background. These are population correlations, around 0.2 to 0.3, which is to say that they are useful for describing groups and weak for predicting any one person, in the way that IQ is.
Can it be changed on purpose
If roles change traits, interventions might. A meta-analysis of 207 studies published in 2017, mostly clinical trials of psychotherapy that happened to measure personality, found that traits changed by about a third of a standard deviation over an average of six months, with the largest change in emotional stability and the change persisting at follow-up. Studies in which people set goals to change a trait and were given weekly challenges have found smaller but real effects over a few months, in the direction people wanted, mostly for people who did the challenges. Nobody has turned an introvert into an extravert in a laboratory, and nobody has tried; the finding is that traits respond to sustained effort by a modest amount, which is what the natural changes of adult life would lead you to expect.
What to tell a 20-year-old
The honest summary is this. Your rank among your peers on each of the five traits is fairly stable and becomes more so, but it is not fixed, and it is least fixed now. Over the next twenty years you will probably become more conscientious, more agreeable, and calmer, whether or not you try, and faster if your life gives you responsibilities. If you want to change a trait, it responds to effort by a modest amount over months, and the research suggests that acting the trait, repeatedly and in real situations, is what moves the score. James's plaster was a good metaphor for his own time, when most of the data came from his own introspection. With 150 studies and a hundred thousand participants, clay that slowly firms is closer.


