I recently stumbled across a study titled Personality Stability From Age 14 to Age 77 Years by Harris et al.. The study examined the personality traits of 1208 14-year olds in 1947. 63 years later, they followed up with the participants who agreed (n = 174). Their interesting result: The authors found virtually no correlation between most of the six character traits respondents were rated on. After correcting for various confounders they find that only ‘Stability of Mood’ remained somewhat stable over a lifetime. ‘Conscientiousness’ was almost significantly correlated, but not quite.
I found this somewhat surprising, as my intuition always was that character traits are somewhat stable over the course of a person’s life. Other studies indeed find more stability of character traits over time, but a in general stability seems to decline over time. Parnas et al. (how can a paper that was published in 1983 still be behind a paywall???) found e.g. character neurosis to be relatively stable when looking at youth from age 15 to age 25). Hampson and Goldberg followed school children in Hawaii and reexamined 799 of them 40 years later. They found modest stability for some character traits, in particular Conscientiousness and Extraversion. Soldz and Vaillant followed Harvard alumni over 45 years and found significant (albeit not very large) correlations for Neuroticism, Extraversion, and Openness.
What to make of this?
Probably not all too much. It is of course a consoling thought that people can change over time. Quartz jubilates and titles an article written on the Harris et al. study: “You’re a completely different person at 14 and 77, the longest-running personality study ever has found“. And I wouldn’t even say that is false per se. I am simply hesitant we can conclude that from the evidence the study provides.
Even if personality traits were one hundred percent completely stable over the course of a lifetime: how likely would it be for a study to detect that reliably? There are so many ways in which results could be confounded that it is hard to tell. Most obviously: Harris et al. asked several teachers to rate the personality of their pupils back in 1947. In the follow up 63 years later the participants were asked to rate themselves and nominate a close friend to rate them. This introduces biases and random variation that is hard to recover. Also the words people used to describe themselves may not mean the same after 63 years. Or the concepts behind the character traits have slightly changed. Maybe the same child would have been rated quite differently by their teacher in a modern class room than back in 1947?
So evidence of missing correlation might also be just evidence for how hard it is to conduct proper research. Intuitively it does seem plausible that people change over time. It also makes sense that studies that included children or young participants found less stability than the ones only including adults. I wouldn’t conclude you are a completely different person. But there is definitely room for change (and maybe even improvement?). So let’s see how we turn out to be at age 77.