First off – this aural null stuff was way before my time, and my time was when fixed NDB cards were standard. Never trained on it nor used it; this is just what I recall reading.
An NDB aerial on an aircraft was a circular loop in the vertical plane. You will recall seeing an open circular antenna on top of old aeroplanes or ones which have been covered by an aerodynamic housing.
If an NDB was broadcasting, an aircraft in range would be able to receive the signal. Say it is heading 060 and it can hear the NDB. To find its bearing to the NDB, the pilot would alter heading until no signal was detectable (the aural null) and then it would be at right angles to the NDB.
I think the null was received instead of using the strongest level because that 100% signal was too difficult to resolve.
But the null didn’t tell you whether the NDB was on the left or right of the aircraft. So, the pilot needed to do some manoeuvring to resolve the ambiguity. When he did, he or she could then track to the station
In general, I am pretty sure that is correct but I do not mind being corrected.
It was a bit like the start of the DME homing procedure in that you took up a heading, watched the miles and groundspeed (or mentally calculated it), tried another heading etcetera.