It’s very easy to conceptualise methods within psychology as being pretty straightforward- maybe not in terms of their mechanisms or procedures, but generally in terms of what they do. EEG measures electrical activity. fMRI measures blood flow changes in response to tasks or stimuli. TMS disrupts activity and let’s us see what happens when we apply this to different areas of the brain. The actual physics of how they work might be pretty complicated, but hey, the ideas themselves are quite simple, right?
Unfortunately, no. I alluded to this a little in a previous post (the one about interpretations, not the one about how much I love my cat…though I’m sure some vague association can be made), but the way we apply these methods, i.e. the experimental paradigms we use in conjunction with these methods, can drastically alter what they tell us.
This is the topic for the coursework that has been set for our (surprisingly interesting and, I believe, quite underrated) Methods course. It focuses on fMRI and how researchers have applied this tool to find out stuff about the brain, from the most basic of processes in, for example, vision, to the really high level stuff, like what areas of the brain are activated when we make decisions or plan stuff or perform any other executive function. The results of these studies are often hailed as revolutionary, plastered across all popular media outlets, and are supported by other studies replicating the same thing using the exact same method, without consideration of how the paradigm they’ve applied has affected or mediated their results.
The three paradigms this paper we’ve been set focuses on are all quite different, and hence in comparison, their strengths and weaknesses become more apparent. The first is the ‘obvious’ application of fMRI: subtraction. Simply put, subtraction involves subtracting activation of the brain in one condition from activation of the brain in another, and then claiming that the difference between the two shows which areas of the brain were involved. So, say you’re looking at face processing. You show participants neutral stimuli, and then the facial stimuli, and the areas that are activated more by the latter condition are supposedly the areas which are involved in the processing taking place- so, in this instance, most likely the fusiform gyrus.
This all sounds very convincing so far, but things are not as straightforward as they seem. Getting the baseline measure is probably the hardest thing. What is this ‘neutral’ stimulus that you’re using? How can you tell activation is due to the fact that participants are seeing a face; maybe other features of the stimuli are causing this increased activation? Is the brain ever truly at rest in order to obtain this comparable baseline measure? Are the neurons that are are not selective for faces responding equally to both the control and stimulus conditions?
And yet subtraction is a very popular paradigm. It obviously does work to a certain extent, but from the papers I’ve read so far, it seems as though not all researchers really consider the limitations of the paradigm they’re using before rushing ahead and conducting their study. The other two paradigms this coursework is on are adaptation and multivariate pattern analysis, both of which, much like the subtraction paradigm, have their strengths and weaknesses. Of course, there is no ‘perfect’ experiment, no ‘perfect’ way to measure the brain, and we can only try our best. However, very often researchers don’t try their best, and this Methods course is fantastic in that it is teaching us to be critical of studies, especially ones involving fMRI. These are too often praised without evaluation of the methodology, even if the theory behind the practice is laudible.
And anyway, learning about methods is (weirdly) probably one of the most interesting things I’ve done on this course so far…It is truly the basis of the subject of ‘psychology’ as we know it today.
paucis verbis 