This paper presents a pilot study on the differences in evoked EEG coherence measures (computed as zero-lagged cross-correlation, "EC") between chess experts and novices while solving chess problems. One of out earlier investigations (Volke, 1999) had shown that both location and direction of EC changes in task processing (compared with a resting condition) depend on the degree of task performance. As a result we postulated that when solving the same tasks experts and novices would show different EC. The results of the recent study confirmed this supposition. Chess tasks of varying difficulty were presented to 25 chess players of different performance levels. The EEG was recorded from 29 positions (enhanced 10 : 20-system). EC were computed for post-stimulus intervals of 1.6s. A special averaging procedure (considering Fisher's Z transformation) was applied for summarizing the single EC of each type of task. These subject- and task-related EC-values (29 x 28/2 = 406 values per frequency band) were subjected to a MANOVA after a two-staged principal component analysis. Significant differences between experts and novices, compared with the resting condition, were found in the delta frequency range. The relevant cortical areas of experts, defined by a topological analysis of the differentiating principal components, were located rather posterior and more in the light hemisphere, compared with those of the novices. Furthermore, the essential EC of the experts showed higher values in the task situation ("on-coupling"), compared with the resting condition, whereas those of the novices were lower ("off-coupling"). The results are interpreted as a higher task-related functional integration of the cortical areas in experts.
