As far as behavior is concerned, it primarily is a response that an animal comes up with as a result of some stimulus. Neurothologists look at this phenomenon somewhat mechanistically - with an aim to unromanticise (?) it, trying to understand how eactly the central nervous system controls an animal’s response to their environment. We are supposed to concern ourselves with rather simple behaviours - sadly, the conceptual, and the technical difficulties one would have to undergo to deal with something more complex, is quite huge.
What does a neuroethologist need in her bag?
Starting off with neuroethology requires you to be aware of three things, as Zupanc mentions:
Choosing the right level of simplicity - classifying an ethogram into subclasses of behaviours, and then try to operate at a lower level of these integrations - perhaps, trying to find out the structures and the chemicals responsible for such a response. This makes room for a better definition of the said behavior, and then quantify it.
Quantifying behavior - This includes, frequency, duration, and intensity, however, this often is impractical.
Choosing the right model system - Simple, robust, readily accessible (or something you fell in love with). Once this is done, you are good to go!