Chris Fraley
R. Chris Fraley is a personality psychologist at the University of Illinois and the author of the meta-analysis that asked how stable attachment actually is over a lifetime. The answer is less stable than the popular account requires, and the number everybody quotes from his paper is not the number he measured.
Who they are
Fraley's 2002 review gathered the studies that had measured the same people's attachment at two points in time, and modelled what pattern of change best fitted them.
Two findings matter here. The first is the long-range one: the observed correlation between attachment at age one and at age nineteen is modest, and it rests on roughly two hundred people in total, with individual studies disagreeing sharply. The figure most often quoted from this paper is larger, and it is not an observed correlation at all. It is the stability ceiling estimated by the best-fitting model. Quoting a model's asymptote as though it were a measurement is a small error that makes the case for lifelong continuity look considerably better than the data do.
The second finding is quieter and does more damage. Measure the same babies at twelve months and again at eighteen, and the agreement between the two is weak, across nearly nine hundred infants. A measure that agrees with itself only weakly across six months is then asked to carry a prediction across two decades.
Fraley's own text is careful about this. The fit, he writes, is far from perfect. He also developed the dynamic modelling approach that distinguishes between a prototype account, in which an early representation persists and pulls later ones back toward it, and a revisionist account, in which each new relationship overwrites the last. The data do not cleanly choose, which is the honest position and the one this site takes.
He is a co-author of the 2023 preregistered test of the Strange Situation's predictive validity.
Key papers
Key papers
Hudson, N. W., & Fraley, R. C. (2015), Journal of Personality and Social Psychology, 109(3), 490-507. The volitional personality change paper. Nathan Hudson is first author and is credited first here; the full account of the design, the effect sizes and the replication trail sits on his page. This entry exists so that the second author is not silently dropped, which is the failure this site logs most often in the other direction.
He is also the author of widely used self-report attachment measures, and his methodological work on measurement and on open science is part of why this corner of personality research is unusually well behaved. No single citation for that strand is recorded in the vault, and none is invented here.
Key ideas
Prototype versus revisionist. Does an early representation persist, or does each relationship rewrite the file? The data do not settle it.
A model's asymptote is not a measurement, and the difference has been quoted away.
Six-month instability undercuts twenty-year prediction, which is the fact this article is built on.
Concepts on this site
Find them elsewhere
Faculty page and lab page both confirmed at illinois.edu. No social account recorded
Checked 2026-09-11. Handles and affiliations go out of date, so treat this block as a snapshot rather than a standing fact.
Sources
- Fraley, R. C. (2002). Personality and Social Psychology Review, 6(2), 123-151.
- Caldo, P. D., et al. (2023). Development and Psychopathology, 37(1), 147-163.