Clemens' answer is correct although it can be dangerous making assignments based on negative data. If the compound is as drawn then I would expect a good NOE between the methyl signals as Clemens suggests. In the other isomer, you would probably see an NOE to the olefinic proton from the acetyl methyl (although this will be weaker). I'm not sure which couplings Khushwant is saying is indicative (there is only one proton on the olefin). There might be a proton-13C coupling to the acetyl carbonyl which would be different in the two isomers but that is a hard experiment to perform and requires a good model.