TABLE 4.
Magnitude of mean between‐group difference (Cohen's d) in neurocognitive domain functioning before and after adjustment for each domain and intelligence
| Neurocognitive domain | Unadjusted domain group difference | Domain between‐group difference and percentage reduction after using each neurocognitive domain and intelligence as the covariate | ||||||
|---|---|---|---|---|---|---|---|---|
| Speed of processing | Attention/vigilance | Working memory | Reasoning and problem solving | Verbal learning and memory | Visual learning and memory | Current intelligence | ||
| Speed of processing | −1.12*** | −0.75*** (33%) | −0.84*** (25%) | −0.98*** (13%) | −0.96*** (14%) | −0.97*** (13%) | −0.86*** (23%) | |
| Attention/Vigilance | −0.88*** | −0.39 (55%) | −0.59** (33%) | −0.71*** (19%) | −0.77*** (13%) | −0.72*** (18%) | −0.67** (24%) | |
| Working memory | −0.77*** | −0.33 (57%) | −0.42* (45%) | −0.53* (31%) | −0.62** (19%) | −0.60** (22%) | −0.48* a (38%) | |
| Reasoning and problem solving | −0.57** | −0.26 (54%) | −0.25 (56%) | −0.12 (79%) | −0.41* (28%) | −0.42* (26%) | −0.01 (98%) | |
| Verbal learning and memory | −0.53** | −0.03 (94%) | −0.33 (38%) | −0.28 (47%) | −0.37 (30%) | −0.40 (25%) | −0.22 (58%) | |
| Visual learning and memory | −0.53** | −0.11 (79%) | −0.18 (66%) | −0.23 (57%) | −0.37 (30%) | −0.39 (26%) | −0.31 (42%) | |
| Mean percentage reduction in Cohen's d | 68% | 48% | 48% | 25% | 20% | 21% | 47% | |
One‐way analyses of covariance (ANCOVAs) comparing ultra‐high risk individuals (n = 50) and healthy controls (n = 50). Columns indicate neurocognitive variable used as the covariate, rows indicate neurocognitive domain used as the dependent variable. Blank spaces represent analyses where a domain would have been both dependent variable and covariate. Negative Cohen's d value signifies worse performance of the ultra‐high‐risk group. Reduction in d value after adjustment is shown within brackets as rounded percentage.
A significant interaction term (p = .015) between group and current intelligence indicated that the assumption of homogeneity of regression slopes was violated, and this result should therefore be interpreted cautiously.
p ≤ .05, **p ≤ .01, ***p ≤ .001: Significance levels.