Human behavior arises from both unconscious and conscious sources. A simple example is breathing. You don’t usually have to think about inhaling and exhaling, but you can can take control you want. By commanding your attention to your breath, you can hypterventilate or elongate your inhales and exhales. This combination of conscious and unconscious processing is an area of intense research. Even researchers who study exclusively conscious cognitive processes (as I do), have to contend with influences of unconscious processing.
A helpful metaphor for understanding human cognition is to think of processes that fall on a consciousness continuum. The consciousness continuum goes from completely unconscious neural activity to the fully conscious explanations we articulate about why we do things. My studies have touched upon 4 layers: neural processing, eye movements, behavior (i.e. actions you take), and feedback (i.e. explanations you have for those actions). Incidentally things at the bottom of the consciousness continuum happen at a much much faster time scale than things at the top. As such, studying memory has required a different tools for understanding each layer.
My research attempts to track phenomena as they influence one another across the various levels of the consciousness continuum.
A human brain is composed of an estimated 100 trillion neural connections. Unsurprisingly, we don’t have much insight into it's inner workings. Cognitive neuroscientists have developed several different approaches to measuring the activity of a living working brain of a human (fMRI, EEG), but each of these techniques has it's own limitations.
fMRI measures blood flow. It roughly tracks which parts of your brain are currently "thinking", because oxygenated blood flows to regions with a lot of neural activity. fMRI has become hugely popular in the medical community because it was one of the first techniques to be able to measure 3D activity throughout an entire human brain with millimeter-scale resolution. However, blood flow is roughly 10,000 times slower than neural activity (10-15 seconds compared to tens of milliseconds), so we can't get precise timing about when the brain responds to a stimulus.
I've also used EEG (electroencephalography), which measures electrical activity from the brain at millisecond timescales. This technique lets us measure neural activity as it unfolds. EEG comes in a couple forms, and I've used two of them: scalp EEG, which measures from electrodes placed on the scalp, and intracranial EEG, which measures from electrodes that have been surgically implanted deep into the brain.
The most common form, scalp EEG, involves placing what looks like a shower cap on a person's head. While it has precise temporal resolution, it only detects electrical activity that has traveled through the brain up to the scalp. We tend to only see activity from brain regions that are located in superficial layers of the brain. Another major weakness is that we are unable to see exactly where the signals are coming from.
Intracranial EEG, on the other hand, can measure activity from specific regions deep in the brain because electrodes on tiny needles have been impaled deep into a patient's brain. Intracranial EEG is used clinically for epilepsy patients seeking to pinpoint the exact location of their seizures. Patients typically have these electrodes implanted for about a week while doctors attempt to localize their seizures. I work with epilepsy patients while they are undergoing this treatment.
Eye movements occur more rapidly than you might think. You look at 3 to 5 distinct places every second. You are only consciously aware of only a subset of these movements. Different processes (e.g. perception, attention, and memory) can dictate what you look at and when, before you have the chance to make a conscious decision about where to look.
For example, imagine a small child did something wrong -- he broke the cookie jar and crudely hid it behind the couch. You ask, what happened to the cookie jar? He doesn’t say anything, but you know immediately by his gaze where the cookie jar remnants are. His eyes couldn’t help but look at the evidence. I've designed experiments to directly measure this phenomena (see Eye Movements post).
Of course we also move our eyes deliberately. We can choose to look at a painting or search under the couch for the tv remote. Regardless of whether eye movements are intentional or not, they can provide rich information about what a person might be thinking as their thoughts unfold.
A behavior is a specific action that you take. For cognitive neuroscientists, behavior is usually measured by how well you can complete a task (e.g. do you recognize an old photo?). By measuring brain activity while subjects are doing very specific tasks, scientists learn which unconscious neural mechanisms are involved in conscious behaviors.
Behavior, however, occurs on much much slower time scales than unconscious eye movements and neural activity. The outcome of a trial lets us see the conclusion the subject's thought process led them to. It does not let us analyze all of the individual events and subprocesses that led the subject to that specific behavior.
Feedback comes in a few different forms. One way I use feedback is as a follow-up to a subject's behavioral response. After a subject makes an overt memory decision (e.g. "Yes, I remember that old photo"), I will ask the subject how confident they are in their response. It turns out that confidence tracks accuracy pretty well -- the more confident people are the, the more accurate their response tends to be. However, confidence and accuracy can diverge when it comes to false memories. People can be highly confident that something happened when they only imagined it or incorrectly remembered it (just think of that last argument you had with your husband about who took out the trash).
The other way I use feedback is to gain insight into my experimental design. When I’m trying out a new experiment and working out the kinks, I’ll sit down with the participants and ask their thoughts on the task. What strategy did they use? Did it feel too hard? Was there anything peculiar about the stimuli? Getting direct feedback from the participants saves a lot of time when I’m designing new experiments. I don’t want to end up with an experiment with some glaring errors that were readily detected by the participants. Also, people are pretty perceptive, so it’s been useful to take advantage of their thoughts to improve my designs.