On the german Wikipedia page on fMRI I read the following
"There is some critisism of the basic assumptions and possible insights gained from fMRI studies which is based on a constructive component of the visualization of measured data, whereby it is more the ideas and expectations of the researchers than the actual processes that are visualized." (translation by me)
Can these "basic assumptions" and "expectations" be made more precise, and the way how they are reflected in the "constructive component", and how the constructive component in turn is reflected in the visualization (which doesn't necessarily show the averaged but true activities of neurons)?
What do typical "basic assumptions" concern? The processing of single neurons, the inner connectivity and processing of groups of neurons, the collective processing of different groups of neurons, the interwovenness of neurons and blood vessels?
What is a specific typical "basic assumption" in the context of fMRI?
What is a typical "expectation of the researcher"? Finding a correlation between two regions of activity while performing a task?
What is a typical "constructive component"? A method of statistical analysis? Or something else?
How does an inappropriate "constructive component" typically lead to false fMRI visualizations?
Do fMRI researchers regularly check if a BOLD signal really (or most probably) reflects neuronal acitivity at that spot - and don't display it if not?