Okay, I did not expect that passage about NMR in the study guide would have sparked a discussion. I do remember having a rather technical debate with Thushan when we were editing that part.
Here is what little amounts I know about NMR:
The typical (and intuitive) setup is, we apply some constant external magnetic field via a very powerful magnet. This field is constant, and is kept
precisely constant in all spatial dimensions. This is achieved with various smart placements of electromagnets so that we get almost-perfect symmetry everywhere. (There are other measures to overcome even the finest variations, but we won't go into that here.)
This field will separate the spin states of the nuclei (parallel and antiparallel to mag.f) by a certain energy, we then vary the radio frequency such that it matches this energy difference, and promote the parallel state to the antiparallel state. We use a RF coil to emit these frequencies, and the RF coil also doubles as the detector. I'm not too sure how exactly RF coils work, but I do know that it emits and detects the same frequency, and shares the same basic principle as your mobile phone antenna. (This is semi-high level electrical engineering, a bit of LC circuits here and there, I was lazy and didn't take the NMR class). So anyhow, the RF coil can be tuned to give off certain radio frequencies, and when radio energy precisely match the energy difference between spin states (an over simplification of the actual mechanism, but it'll do), the RF coil gives you a response which results in a peak on your graph.
Now, that's the intuitive setup.
HOWEVER, the difficulty comes in tuning the RF coil, which typically only resonate in a narrow range of RF, and is hard to tune correctly. Some instruments do successfully tune the RF coil and obtain NMR spectra this way, but many opt for the easier option of varying the applied magnetic field by varying the current in the electromagnets. Keep in mind that even as we vary the current, the external/applied magnetic field still maintains its spatial uniformity, so as far as the molecules concerned it's still in a uniform magnetic field (that changes strength over time). The theory about how everything else work is roughly the same, except in this case, we vary the magnetic field until the sample's

in spin states resonates with the fixed RF frequency.
Both of these cases (varying magnet/varying RF) are possible. Knowledge of neither are required for VCE. Heck, knowledge of neither are required for even a chemistry major at uni, and remains as an elective for honours/masters students.
So in summary, find yourself a bullshit answer you're happy with to get by VCE in case they ask you "derp what is a chemical shift", otherwise just learn how to read the graphs.
PS holy shit you guys amaze me. When I first learnt about NMR in VCE I was scratching my head going "wtf magnets", not discussing which of the RF or the applied magnetic field was being varied.