Reading focus
Syllabus: How Stable Is Personality?
Compare the age patterns for mean-level change and rank-order stability, attending to measurement intervals and the studies available at each age.
Why you’re reading it. Replace the simple question ‘Does personality change?’ with a qualified account of what changes, when, and for whom the evidence is strongest.
A question to consider. Alex is 35. What can the meta-analysis tell us about change at that age, and why can it not forecast Alex’s trajectory from the average curve?
Why this reading
This synthesis asks how personality stability and change vary with age, trait, measurement, and the interval between observations. It brings mean-level and rank-order evidence into the same account, helping readers evaluate broad developmental claims against the studies that support them.
Before you read
A meta-analysis combines results from multiple studies; it does not follow one cohort from birth to old age. Rank-order correlations describe relative standing, while standardized mean differences describe changes in averages. A moderator is a characteristic associated with differences in results across studies; it is not automatically a causal explanation.
Read for
- Connect the inclusion criteria and measurement intervals with the forms of stability this review can address.
- Compare the age patterns for rank-order stability and mean-level change without treating either as an individual forecast.
- Examine variation across studies and the strength of coverage at different ages before interpreting a fitted lifespan curve.
Consider afterwards
- A headline says personality becomes fixed at age 25. How would you use the distinction between a plateau in rank-order stability and continuing mean-level change to evaluate it?
Editorial context
The evidence supports considerable stability alongside change; a plateau in one stability statistic is not an age at which development stops. Studies from Northern Europe and North America predominate, and evidence is thinner at the oldest ages. The review’s eligibility rules exclude clinical and intervention samples. Its averages therefore cannot directly estimate what a particular intervention will do, or establish a universal developmental timetable.