Sometimes analysing neuronal data is easy, but sometimes it isn’t. One thing that frustrates me almost daily is when people use the difficulty or complexity of some kind of analysis as a justification for not doing it. Our approach to analysis needs to be based on what makes the most sense based on the question we are asking, not what is the easiest or what we have experience doing in the past. Neuroscience is unbelievably complicated; it would be silly to think that we can answer all questions by just counting spikes in arbitrary windows.
If the question is – which stimulus parameters alter the strength of neuronal responses, then the classical approaches of counting spikes in windows usually work perfectly. However if the question is – What are the responses of this neuron used for, or how might responses of this neuron be interpreted by downstream neurons, spike rates over a window make almost no sense at all. Neuronal communication cannot ever be broken down to spike rates over arbitrary windows, even in neurons that use a rate code (which many would argue don’t even exist). This spike rate will be correlated with the volume of information being transferred, but it gives you absolutely no insight into the real question, what does that information tell us (or more importantly, tell the downstream neuron)?
I personally think so many neuroscience questions are held back by analysis which is simple, but completely misses the main point. So many of us chose to be neuroscientists because neuroscience is so complicated and amazing, but then run and hide when our data needs complex analysis.