Most studies I see compare firing rate with some observed variable (stimuli or behavior). However, they treat all spikes the same. But neurons sometimes release different neurotransmitters depending on the situation, meaning not all spikes are the same. Does this affect our interpretation of these studies? Are there models that account for this? Do spikes vary in any other way?
For example, knowing when a neuron is acting as excitatory versus inhibitory may vastly change our interpretation of it's role.
Additionally this may become even more of an issue when we attempt to model or predict from multiple channels as we are potentially missing out on a crucial piece of information. Without that, do we even have the ability to fully understand the area?
On a slightly different note: Are there other aspects of a neuron output that are important in it's computational role in the circuit? Why don't we take into account it's amplitude over-time for example?
My primary worry is that we are averaging away, or outright throwing away, a lot of information that which could be insightful.