A critical review of EEG coherence studies of hemisphere function.

Studies of hemisphere function involving the use of coherence analysis are reviewed. Despite the fact that many effects related to subject group and type of task have been reported, few seem to be reliable across studies. The assumptions which seem to underlie the use of coherence analysis are made explicit and discussed. Most investigators consider coherence levels to reflect certain 'baseline' patterns, possibly related to structural connectivity, upon which task and/or strategy effects may be superimposed. Use of a cephalic reference renders coherence effects virtually uninterpretable for a variety of reasons which are fully discussed.