Fig. 6.

A: Illustration of frontal alpha asymmetry (FAA) means and 95% error bars (representing standard error of the mean) for controls and depressed patients, separately for the three different EEG reference schemes. B: Similar to A, except for the calculation of FAA without dividing by the sum of F4 and F3. Note the differences based on EEG reference montage, but also that none of these methodological changes changed the overall MDD-control contrast to a significant difference, illustrating that these methodological aspects could yield different outcomes, but do not explain the lack of ‘diagnostic’ effect of FAA in this large sample.