Counter evidence with counterevidence; it could be a good assumption. Even if someone's personality is not static, it doesn't mean that the variation is large enough to be significant, either.