so matt did look at pre stim theta power and it did not show anything
the prediction would have been that both alpha power and phase mattered, but that is not what we found
this disconnect between power and phase is not uncommon, in fact you would typically analyze the two separately.
the exception to that is the fact that there is some intrinsic link between the two. namely that with lower power you have less oscillations and thus your phase measures will be unclear.
past work on pre stim alpha phase vs power has shown that phase is typically more predictive of performance then power
the idea here is that phase matters most when you have strong alpha alpha power states; and that it matters less with weaker alpha power states
this goes back to the idea of alpha as inhibitory vs being about attention
if you know a stimulus is going to occur or you entrain to something, you can have strong alpha power where the phase aligns maximally for processing the stim
that is to say, you have the ability to inhibit cortical excitability with less risk of missing something
in situations where stim onset is ambiguous, you might go into a low alpha power state so as to insure that you have maximal excitability to catch the stim.
the big idea here is that the inhibitory function of alpha is not continuous,. instead it is pulsed (for example, inhibition happening at 10 times per second), being maximal at at specific phase of alpha.
so you are either getting inhibition or no inhibition, and oscillating between the two states
when in low alpha power; you get basically no inhibition.