I don't understand proton NMR. Help me understand...

So I understand the theory behind proton NMR, singlet, doublet etc... D shows no peaks, and so on.

I just don't know how to apply the knowledge to exam questions, it's losing me at least 8 marks most occasions.

For instance, when looking at the data sheet http://www.ocr.org.uk/Images/74947-datasheet.pdf and go to 2.5ppm, I don't know what to pick out of the 4 possible structures! How do I go about learning how to differentiate between them?

In any question you'll have other information about the molecule than just NMR usually some combination of IR, Mass Spec and an empirical formula or some observations for reactions of the compound which can give you and idea of the structure.

I do find that other pieces of info like Mass Spec definitely help, but there's one example where they give the carbon NMR and I just don't get it.
Q4d for instance http://pmt.physicsandmathstutor.com/download/Chemistry/A-level/Past-Papers/OCR-Old/Unit-4/June%202015%20QP%20-%20Unit%204%20OCR%20Chemistry%20A-level.pdf is rather difficult, and so I don't know which compound to pick from the data sheet for each ppm value.

The proton NMR gives you one of the R groups for free. There's a singlet with integration trace 3 at about 1.8ppm, which they tell you corresponds to a HC-C=N environment.

The other is just noting there are 7 aromatic protons (the peaks at about 7ppm) and there's a CH₂CH₃ from the splitting of the remaining peaks. There's only one way to arrange a CH₂CH₃ and extra phenyl group that produces the correct splitting.

The C NMR lets you place the ethyl substituent at the right position on the phenyl.

Ooooh that would make sense! I'm just not sure on how you could tell the CH2CH3 from the other peaks?

Triplet (3H) and a quartet (2H). Classic pattern for an ethyl group attached to something the protons can't couple with. The CH₃ protons couple with the 2 CH₂ protons giving a triplet, the CH₂ protons couple with the 3 CH₃ protons giving a quartet.

So with the quartet (2H) would you not have to compare it to the data sheet, or do you just use the integration trace?

You should always comment on the chemical shift, the splitting and the integration trace.

However often you can assign the peaks based on just some of the information and explain everything else after the fact.