Figure 2.

Correlation coefficients of true and predicted executive function variables computed with classical feature set. Executive function variables were predicted based on the sum scores of all 3 semantic verbal fluency tests as well as the total sum score across these 3 tests, which led to a total number of four verbal fluency features. Results shown in this table illustrate executive function variables which could be predicted with p < 0.01 from verbal fluency data; Colour groups indicate EF domains and gradients denote different EF tests within this EF domain; MACK Mackworth Clock Test; NVLT non-verbal learning test; SPM Raven’s Standard Progressive Matrices; STROOP Stroop Test; WAF-G Divided attention; WAF-R Spatial attention.