You have already assumed correctly, here are some details on the principle.
Loop antenna principle
The loop antenna is one of the antennas that work using the magnetic field rather than the electric field of the wave. Than means it can be seen as a coil and modeled using Faraday's law of induction. As a NDB transmits a vertically polarized wave, the magnetic field oscillates in the horizontal plan and the ADF loop antenna must be used horizontally.
The working principle of a loop antenna is the following:
Magnetic field crossing a loop antenna in its far field, view from above, source
The peak reception is obtained when the plan of the loop is aligned with the direction of the NDB. The electromagnetic field hits each side of the loop with a different phase. This induces a potential difference and a current circulates accordingly. The minimum or null signal is received when the loop is facing the transmitter. In this configuration, the wave hits all the points of the loop with the same phase and no voltage is created.
As a side note, the size of the loop influences antenna efficiency. The potential difference is maximum when sides length is 1/2 wavelength (for a NDB this is about 500 m). On a practical point of view, large loops have large wire impedance with cancels partly this efficiency increase, so a reasonable compromise has to be found.
If we represent the relative efficiency in each direction, we obtain this type of pattern (assuming the transmitter is up):
Rate of variation of gain
A full radiating pattern informs about the actual gain of the antenna for a given frequency:
The gain is provided in decibels, which is a coefficient. The reference of 0dB is the direction of the maximum gain and negative values indicates a loss in the reception compared to the reference.
When looking at the indications in green, The dip in received energy (the null) is narrower than the peak. This tells us about the rate of variation of the sensitivity:
Signal is maximum at 0°, but varies only by less than 2 dB on a range of 60° around the maximum.
Signal varies by more than 30 dB for only 30° around the minimum/null.
From that it's possible to say that it will be more accurate to determine the direction of the null rather than the direction of the peak (which is the direction of the transmitter). This is the case of many types of antenna: Nulls are more pronounced than maximums.
Loop antenna with sense antenna
There is a problem with the loop antenna: It has a symmetrical radiation pattern, and there are two possible directions for the transmitter. So usually the pattern is modified to a cardioid by using an additional element, the sense antenna, which is a vertical omnidirectional antenna:
One lobe of the loop antenna adds with the sense antenna, the other is subtracted creating a dissymmetry (gain increased forward, decreased rearward). The resulting pattern:
The ambiguity isn't possible anymore. Note that the null is now opposite to the maximum, while they where at 90° before adding the sense element. When the null direction has been determined, the NDB direction is the exact opposite.
Modern ADF antenna
On modern ADF, the antenna doesn't rotate to determine the direction of the NDB. Instead the antenna system is a combination of multiple fixed antennas electronically commutated. By selecting different associations, an appropriate radiation pattern is formed and rotated.
A practical realization: