tag:blogger.com,1999:blog-9192597712746432631.post1149474041746175163..comments2024-03-09T09:06:35.288+00:00Comments on Notes from Two Scientific Psychologists: On why fMRI is bullshit, even when you're doing it rightAndrewhttp://www.blogger.com/profile/16732977871048876430noreply@blogger.comBlogger6125tag:blogger.com,1999:blog-9192597712746432631.post-16560598862194328282010-06-15T21:06:48.820+01:002010-06-15T21:06:48.820+01:00By 'ask the right question' I mean he pres...By 'ask the right question' I mean he presented the nervous system with ecologically valid information that he had complete control over. He didn't just have a video of some driving.<br /><br />And I am indeed still saying 'who cares' about this. I found that even when done right it still wasn't satisfying and that was interesting.Andrewhttps://www.blogger.com/profile/16732977871048876430noreply@blogger.comtag:blogger.com,1999:blog-9192597712746432631.post-75964113638510480152010-06-15T19:01:40.326+01:002010-06-15T19:01:40.326+01:00But does John ask the right question? I've not...But does John ask the right question? I've not actually read the paper, but given the title, I assume it is determining the neural correlates of steering yourself (or seeing yourself steered about). Determining these neuronal mechanisms for the sake of determining these neuronal mechanisms seems to be the sort of (very expensive) who cares study you were unhappy with, irrespective of how well controlled the stimuli were.Gavin Buckinghamhttps://www.blogger.com/profile/08539613027114375642noreply@blogger.comtag:blogger.com,1999:blog-9192597712746432631.post-48864834705979909082010-06-15T07:49:48.445+01:002010-06-15T07:49:48.445+01:00You'll notice I followed that up right away wi...You'll notice I followed that up right away with a qualifier :)Andrewhttps://www.blogger.com/profile/16732977871048876430noreply@blogger.comtag:blogger.com,1999:blog-9192597712746432631.post-62013516751241852522010-06-15T01:07:38.293+01:002010-06-15T01:07:38.293+01:00WOW! Andrew actually said "I might not actual...WOW! Andrew actually said "I might not actually be right" -- that's about as close as he ever gets to admitting that he was wrong. :)Mattnoreply@blogger.comtag:blogger.com,1999:blog-9192597712746432631.post-39138936544490222712010-06-14T18:35:45.975+01:002010-06-14T18:35:45.975+01:00The answer to part of this is the previous rant: I...The answer to part of this is the previous rant: I don't actually think fMRI can give us even the types of answers you suggest, because I don't think in general the question is posed correctly. <br /><br />Second, I think ecological types get a bad rap on the brain. The reason we don't care about the brain is not that we don't think it's interesting; it's that we don't think that studying what the brain does makes sense without realising perception begins in the world, not at the eye. This makes neuroscience hard, actually, and beyond most current methods (you can't move in an fMRI, for example). <br /><br />Third, I do take the point that if you could ask the right question (as John does) you might then get an answer of interest to someone interested in what the brain does. But you still end up with the claim that a certain piece of the brain does a certain thing, when actually what happens is that most of the brain does certain things at certain times under certain conditions - there's no way to get back to the system level that, I think, is actually going to be responsible for doing whatever it is that the brain is doing.<br /><br />Let me put it this way: even John started his talk by acknowledging that in this task, the entire brain goes off. Say Parts A & B go off in Task 1, while A, B and C go off in Task 2. An fMRI study would conclude that C does whatever is different in Task 2 as compared to Task 1; but what if B actually does different things when working with A than with A & C? I think this is highly likely to be the kind of thing that goes on; the brain is incredibly dynamic and changing at the axon level on millisecond timescales. B isn't a circuit that has been formed by experience; B is a collection of neurons that get recruited into various systems at various times. <br /><br />I just think that fMRI is the wrong question. This post was just me thinking that even when done properly, according to the rules and with one eye on the information, you don't learn much.<br /><br />Keep defending imaging though; I might not actually be right (although I obviously don't really think that's true :)Andrewhttps://www.blogger.com/profile/16732977871048876430noreply@blogger.comtag:blogger.com,1999:blog-9192597712746432631.post-8019562543673560112010-06-14T18:05:33.580+01:002010-06-14T18:05:33.580+01:00I will try and once again leap to the defense of n...I will try and once again leap to the defense of neuroimaging...<br /><br />I am in general agreement with your stand on fMRI as a supplement to behavior. We found region x in the brain that is interested in the crucial manipulation in our task (say, direction of optic flow). So, this brain region corresponds with the change in behavior that we see with our manipulation. Sure, whoop-de-do. Some part of the brain is involved in this perceptual discrimination. Hey, perhaps its even interesting that this part of the brain is not in the back of the brain, where visual processes are thought to happen. <br /><br />Perhaps there are some clever conclusions that can be drawn from combo behav and neuroimaging studies - task A and task B are, on face value, quite similar. However, they recruit vastly different neural structures to be completed (the dorsal/ventral stream dissociations are a pretty good example of this).<br /><br /><br />But still, not really a huge bang for your buck in terms of our understanding of what makes us tick, in terms of inputs and outputs. Nothing that couldn't be figured out with clever behav experiments.<br /><br />But this is solely from the behavioral point of view - as an ecological psychologist I can see why behavior is all that interests you. Behavior is really all that matters at the end of the day - the only 'outcome'.<br /><br />But a whole other class of researchers cares little about behavior directly - they only want to understand how the brain, as an organ 'works'. Investigating whether brain region X, which previous research has shown to be interested in a certain task (motion perception), also plays a role in more complex variants of the task (optic flow in depth). If so, then the 'purpose' of that region gets redefined. New light is shed upon the previously puzzling symptoms of patients with lesions including that particular part of the brain. Clever behav experiments are designed to pick seemingly identical behaviors apart. This sort of thing can also be accomplished by fairly brutal monkey work. But I suspect that monkeys are an even more expensive way to go about addressing your questions, and certainly not an alternative for those weak-of-stomach.<br /><br />None of this is, I suspect, the point you were making - you study a brand of psychology that, almost by definition, cares nothing about the underlying brain processes. <br /><br />And of course, this is your blog and you are at liberty to say whatever you feel like. But do you really believe that functional imaging is, when used appropriately, an extremely powerful counterpoint to understanding brain and behaviour?Gavin Buckinghamhttps://www.blogger.com/profile/08539613027114375642noreply@blogger.com