Do Big Five Traits Change Less Than We Think? The 2026 Study

What did people think would stay the same, and what did researchers compare?

The 2026 study found a mismatch between perceived and measured change. In its first study, 887 adults in a nationally representative US sample rated Big Five traits as less changeable over the lifespan than nearly all the other characteristics in the comparison. The Big Five are broad dimensions used to describe recurring differences in openness, conscientiousness, extraversion, agreeableness, and neuroticism. The result concerns what this sample believed about change, not a record of how each participant’s own traits had moved.

A second analysis brought together eight longitudinal datasets from the United States, Germany, Australia, and the Netherlands, totaling 166,971 people. It compared standardized yearly and lifespan mean-level change in Big Five traits with change in life satisfaction, subjective health, income, and self-esteem. Mean-level change describes whether the average score in a group rises or falls. The paper reports that Big Five change was comparable to, and often greater than, change in those comparison domains.

That comparison supports a bounded conclusion: people may regard personality as unusually fixed relative to several other parts of life, while population records show a less exceptional degree of average change. It does not establish why people perceive stability, and the two studies do not directly compare each respondent’s belief with that same person’s repeated trait scores. The comparison also depends on which traits, life domains, samples, time spans, and measures were included. So the study challenges a blanket assumption of exceptional fixedness; it does not show that personality is generally fleeting.

The word “standardized” matters here. Researchers put changes measured on different scales into a common form so they can compare their relative size. That makes a cross-domain comparison possible, but it does not mean the original measures asked identical questions or captured equally visible parts of life. Nor does a similar standardized amount mean a change in income feels like a change in a trait. The paper compares the size of average movement, not the meaning or personal importance of each kind of change.

Sources: Personality traits are less stable than people think

How can traits be stable and still change on average?

The answer depends on what “stable” means. Mean-level change tracks movement in a group’s average. Rank-order stability tracks whether people who scored relatively higher or lower than others tend to keep that relative position later. These are different questions, so evidence of average change does not mean that everyone changes in the same direction or that people’s relative ordering disappears.

Imagine a group in which many people become somewhat more comfortable speaking up over time. The group average could rise even if the people who were more talkative at the start still tend to be more talkative than their peers. This is an illustration of how the two measures can coexist, not a result reported by the 2026 study. A broad impression that someone remains recognizably themselves can therefore be compatible with movement in trait levels or with a change in where that person stands relative to others.

A 2022 meta-analysis of longitudinal studies treats rank-order stability and mean-level change as separate but complementary measures. It describes rank-order stability through test-retest correlations and mean-level change through average increases or decreases. Its synthesis found moderate rank-order stability alongside evidence of mean-level development. The authors also note variation by age, time between assessments, and differences in the underlying studies. This longer research record makes “traits are simply unstable” too strong a reading of the new comparison.

A smaller, focused example points in the same direction. In a study of 251 people followed from college at about age 18 to midlife at about age 40, Big Five traits were moderately to highly stable over 20 years, while average trait levels increased. The sample is specific and does not replicate the 2026 comparison with other life domains. It does show why continuity and change need not cancel each other out: people can preserve relative patterns while average levels move.

These measures also differ from the everyday question “Am I still myself?” A correlation in relative ordering is not a certificate of sameness, and a group mean moving is not evidence that a particular person has changed. The meta-analysis notes that estimates vary with the interval between assessments and the ages represented. A comparison over several years can therefore answer a different question from one spanning early adulthood to midlife. Time scale belongs in any claim about stability.

Sources: Personality traits are less stable than people think; Personality stability and change: A meta-analysis of longitudinal studies; Stability and Change in Personality Traits and Major Life Goals From College to Midlife

How can you use the finding without treating it as a personal verdict?

Use the study to make change a reasonable possibility, not to declare that you or someone else has changed. A group average cannot settle one person’s history. To reflect on a particular tendency, name it precisely, choose two periods to compare, and look for repeated behavior in roughly comparable settings. Ask what changed in the person, what changed around them, and what evidence would make you reconsider the first explanation.

For example, taking more initiative in a new role could reflect a real shift in a person’s recurring tendency, a role that now requires initiative, or both. One visible behavior is not enough to decide among these explanations. The 2026 paper does not test that individual case; it gives a population comparison that helps loosen the assumption that trait levels must remain fixed. If the settings differ sharply, treat the comparison as tentative. A different opportunity to act can look like a different tendency, while repeated examples across settings make a change interpretation more informative.

The verdict is that people seem to underestimate how much Big Five averages can change relative to several measured life domains. The important caveat is that this is a comparison of group-level patterns, with evidence drawn from particular measures, countries, and periods; it is not a forecast for every person or trait. To test your own impression, ask someone who knew you in both periods: “What is one specific way I seem to have changed, and one pattern you still recognize?” Their examples can give you something concrete to examine, while differences in memory or context remain possible explanations.

Make the question easier to answer by choosing a behavior rather than asking for a general character judgment. “Have I become more likely to organize plans before a group trip?” points to an observable action; “Am I more conscientious now?” invites a broad label. Then check whether the pattern appears more than once and whether the same opportunities existed in both periods. This small discipline cannot turn memory into a formal measurement, but it reduces the chance that one unusually demanding week stands in for a lasting tendency.

Sources: Personality traits are less stable than people think; Personality stability and change: A meta-analysis of longitudinal studies; Stability and Change in Personality Traits and Major Life Goals From College to Midlife

Questions readers ask