A suggested topic is whether you think the psychological experiments covered today (visual world paradigm) validly test the way we comprehend language

A suggested topic is whether you think the psychological experiments covered today (visual world paradigm) validly test the way we comprehend language.
The visual world paradigm does not validly test the way we comprehend language but it is probably the only way now.
The visual world paradigm relies on the premise that people will predict the next word they hear and match it with an image on a field of images, causing eye movements. These eye movements are then recorded. The time taken for the eye movements is then compared among different groups of people.
Yet these results seem to be very dependent on the people involved. For example, if an image of a cow and a fly is produced and the word is ‘animal’ then the people involved might look at the fly instead of the cow, since flies are animals too. People might also not respond and produce the eye movements. In order to avoid these problems many people have to be tested.
A large sample size is required then, because of the above problems. Yet the standard way to get volunteers is to recruit psychology students – hardly the average person off the street. That already predisposes the subjects to certain biases which certainly could not be average.
Yet the alternatives are not very palatable either. The only viable alternative seems to be to scan the brain while the experiments are being conducted yet that seems to be expensive and also difficult to do as a large sample size is required again. This brings back the problem of the sample size not being representative of humans again. Also another problem is that the interpretation of brain signals can be flawed due to insufficient technology or understanding of the human brain.
Therefore I conclude that the visual world paradigm is flawed, but probably the best choice for conducting such psychological experiments today