Abstract
This study evaluated the feasibility of including a computerized dynamic test of analogical reasoning in standard neuropsychological assessments in a heterogeneous psychiatric population. The participants were 40 adult patients (M(age) = 33.15 ± 12.27, range 19-68; 60% male) enrolled in specialized mental health and forensic care programs in The Netherlands, who were randomly assigned into either a training, a passive, or a control group. A pretest-training-posttest paradigm was used for the training group, and the dynamic test consisted of 26 items of the A:B::C:? type. In terms of practical use, it was found that the administration time varied largely, and 22% of the data was lost due to drop out or technical malfunctions. Test-retest reliability was acceptable for the training group (r = 0.61) and good for the practice and control groups (resp. r = 0.88 and 0.80). A statistical trend was observed for the training vs. practice group (Z = -1.598, p = 0.055), but not for the training vs. control group (Z = -0.839, p = 0.201). It was concluded that an indication of training effectiveness was found; however, in this clinical sample, the applicability of the current form of the dynamic test is still limited. Several modification options are discussed.