Table 1.

Shows exemplar data from one subject visit. Each row summarizes a group of MEG sensors (LT—left temporal, LC—left central, LF—left frontal, LP—left parietal, LO—left occipital, RT—right temporal, etc). The second column shows the change in mean peak frequency in the 2–7 Hz band between the average peak frequency in runs A and B as compared to C (this corresponds to the assumption ‘B sham’); and the third column shows the average peak frequency in runs A and C as compared to run B or ‘C sham’. For example, looking at MRP we see that the mean peak frequency over these channels has increased when we assume that run C is the sham. The last row shows the average frequency change over all channel groups under the two hypotheses. In this case therefore, as we assumed that the pT device would give rise to a frequency increase, we would predict that C was the sham condition.

Chan group	Assume B sham frequency change (Hz)	Assume C sham frequency change (Hz)
MLF	−0.5	1.4
MRF	−0.4	0.2
MLT	−0.3	0.5
MLC	−1.2	1.7
MRC	−1.4	2.1
MRT	−0.4	−0.1
MLP	−1.7	2.2
MRP	−2.5	3.0
MLO	−0.3	−0.4
MRO	−2.9	0.1
Mean all groups	−1.16	1.07