Table 4.
Neuropsychological tests of working memory and reasoning.
| Cognitive Function | Cognitive Measure |
L DLPFC | R DLPFC | Focal NonDLPFC |
NO LESION |
Kruskal- Wallis χ2 |
Kruskal- Wallisp value |
Significant betweengroup differences |
|---|---|---|---|---|---|---|---|---|
| Maintenance | Digit Span Forward | 6.00 (1.41) | 5.89 (1.05) | 6.40 (1.19) | 6.68 (1.22) | 4.83 | 0.19 | none |
| Maintenance | Spatial Span Forward | 8.57 (2.94) | 9.11 (3.89) | 9.89 (2.56) | 10.22 (2.7) | 3.05 | 0.39 | none |
| Monitoring: No Cognitive Load | Zero-Back Errors | 1.00 (0.71) | 2.56 (3.47) | 1.16 (1.89) | 1.08 (2.33) | 3.09 | 0.38 | none |
| Monitoring: Low Cognitive Load | One-Back Errors | 3.40 (2.97) | 3.67 (3.24) | 3.00 (2.75) | 2.57 (2.48) | 1.26 | 0.74 | none |
| Monitoring: Medium Cognitive Load | Two-Back Errors | 5.20 (1.92) | 4.89 (2.37) | 5.26 (2.81) | 3.85 (2.39) | 6.66 | 0.08 | Focal NonDLPFC > NO LESION* |
| Monitoring: High Cognitive Load | Three-Back Errors | 6.00 (2.74) | 5.67 (3.61) | 6.42 (2.65) | 4.68(2.40) | 5.27 | 0.15 | Focal NonDLPFC > NO LESION** |
| Manipulation | Letter-Number Sequencing | 8.14 (2.91) | 9.44 (3.68) | 9.95 (2.95) | 11.04 (2.66) | 7.73 | 0.05 | L DLPFC < NO LESION** |
| Manipulation | Digit Span Backward | 4.14 (1.22) | 4.67 (1.41) | 4.75 (1.21) | 4.96 (1.41) | 2.49 | 0.48 | none |
| Manipulation | Spatial Span Backward | 8.86 (3.81) | 11.33 (2.96) | 11.28 (2.49) | 12.02 (3.12) | 6.02 | 0.11 | L DLPFC < NO LESION* |
| L DLPFC < Focal NonDLPFC* | ||||||||
| Reasoning | Arithmetic | 8.71 (3.82) | 8.22 (3.11) | 10.40 (2.30) | 11.00 (2.25) | 8.70 | 0.03 | R DLPFC < NO LESION** |
| Reasoning | Matrix Reasoning | 10.57 (3.78) | 9.44 (2.79) | 11.89 (2.47) | 12.28 (2.94) | 7.94 | 0.04 | R DLPFC < NO LESION** |
| R DLPFC < Focal NonDLPFC* |
Means are presented with standard deviations in parentheses. Because the assumptions of parametric statistics were not satisfied (Supplemental Table 1, Supplemental Fig. 1 – 22), non-parametric statistics were used to test for group effects and for the pairwise comparisons. Significant between-group differences were determined with the Mann-Whitney U Test. Neuroscience evidence supporting the involvement of the dlPFC in working memory motivated the assessment of between-group differences when the Kruskal-Wallis test did not reach significance (see Introduction).
*
p < .05;
**
p < .01.