As someone who uses only nodal analysis for all problems, I may not be a good source of unbiased advice. However, back when I was a first-year student it was difficult to grok nodal analysis, so all problems were solved with mesh currents. But now I can just look at a circuit and write the node equations without much thought, and with 99.99%+ accuracy. So here is the advice I would have liked to have given myself: Forget what the professor said and find/devise a method for writing node equations that is intuitive to yourself. A non-intuitive method is the mind killer; it leads to conceptual errors and simple mistakes. When analyzing circuits with active devices, usually it is the signal voltages that are important (not currents) so calculate with node voltages. Test points in a production circuit are there for measuring voltages. Soon you'll be getting a symbolic algebra package (like Maple), so the most important analysis criteria is to get the system of equations correct the first time - practice looking at a schematic and typing the circuit equations directly into your algebra package until the method is flawless.
