help needed please x Watch
If you're feeling swamped, I find it helpful to:
a) Start with what you do know
b) Get that organised
I'm going to imagine there are some key areas of cognition that computer modelling is used for and some key methods of doing that modelling. Get them, write them down as headings and organise the bits you do understand under them. Then pick one and read about it - rather than the whole topic. Obviously, some bits will cross over.
As you read more, make notes and continue to organise under your headings, you'll make subheadings (pro's, cons, evidence base etc.) and the essay will start to form.
No essay encompasses everything, so if an area is totally baffling you, just ignored it - unless it is integral to the topic!
No idea if that will help or not... Good luck!
WOW THAT WAS LONG I'M SORRY THIS IS JUST WORD VOMIT FROM THE TOP OF MY HEAD
• Basically, computer models are used to simulate cognitive processes - so the computer is given input (a way to process the information) and performs internal operations to create the behaviour (which may resemble behaviour at the systems level, cellular/molecular level etc)
• So a simple computational change can yield major behavioural changes in two different creatures (with no difference in how the models process information, only in patterns of connectivity) (e.g. Braitenberg, 1984)
• The architecture of computer models resembles (at least on the surface) the nervous system (e.g. processing is distributed across many units; the contribution of any one particular unit may be small in relation to the model's total output; computations are simulated to occur in parallel to each other)
• These models can then lead to testable predictions (e.g. have a look at Desmond & Moore (1991), which confirmed prediction of a model). So basically, computer simulations can be a useful way to develop theory, which can then aid scientists/ psychologists in designing experiments, making predictions, testing theories and interpreting results.
Limitations of computer models, however, can be:
- Over-simplification of how the nervous system is modelled, which means they are limited in scope. E.g. it's not always possible to confirm whether they respond to single neurons or groups or neurons
- ‘Catastrophic interference’ – loss of information when new material is presented
- Most computer models are restricted to quite narrow problems (e.g. the Stroop effect) – so often are less useful in actually generating new predictions
Hope that helps.