EV-0046
Traits do change, and roughly on the timescale of this program
Roberts, B. W., Luo, J., Briley, D. A., Chow, P. I., Su, R., & Hill, P. L. (2017). Psychological Bulletin, 143(2), 117-141 · Source
What this entitles us to say
A systematic review of 207 studies finding that personality trait measures changed measurably through intervention - around d = .37 over an average of 24 weeks - with the changes replicating across experimental and non-experimental designs, holding for non-clinical interventions, and persisting in longitudinal follow-up.
What it found
207 studies tracking changes in personality trait measures during intervention, including true experiments and pre-post designs. Interventions were associated with marked change over an average of 24 weeks, around d = .37. The increases replicated across experimental and non-experimental designs, held for non-clinical interventions, and persisted in longitudinal follow-ups beyond the end of the intervention. Emotional stability changed most, then extraversion.
Why it matters here
This is the foundation of A16, and it settles the question the whole program rests on: can a person actually change one of these?
The answer the research supports is yes, deliberately, by a moderate amount, over months rather than days - and it holds after the intervention stops. That is a far more defensible claim than the leadership-industry version, and it is enough. A program does not need to promise transformation. It needs to be able to say that deliberate, sustained work moves the needle and the movement lasts.
The 24-week average is worth noticing rather than overusing. The leader program is 26 weeks. That is a coincidence of timescale rather than a validation of the design, but it does mean the program is built at roughly the duration at which measured change has been observed - and it is an argument against ever compressing the leader track for convenience.
It also disciplines what the re-assessment at three and six months may claim. Movement of this size is detectable in aggregate across a cohort. It is not reliably detectable in one person on one trait, and the movement report should be written knowing that.
The limits
The primary focus is clinical intervention - therapy - and the largest changes were in emotional stability, followed by extraversion. Neither is the part of personality this program works on. The authors report that results replicated for non-clinical interventions, but the bulk of the evidence is clinical and that must be said out loud. It also measures self-reported trait scores, which can move because the person changed or because how they see themselves changed. And personality traits are not the same as the twelve character traits in this model - the mapping is ours. Verified to abstract level.
What we do not say
"Research proves character can be changed by training." The bulk of that evidence is clinical, the traits measured are Big Five personality rather than character, and the mapping to our twelve is ours. Say what it is.