highly significantp < 0.001
Our results are drawn from a large sample ( N = 2,761) with high between-subject variability, circumstances that translate into very small but nonetheless highly significant correlations between AIM score change and CaGI-C (r Spearman ~ 0.10, p < 0.001).