Modelling

Methods

Start date:

Commentary

Modelling – theoretical, cognitive, neurological. A psychological model is a theoretical construct which aims to help psychologists understand psychological phenomena through simplification. This is done by developing a representation that aims to represent important aspects of the phenomenon through reducing it to its essential features. By ignoring the less central aspects of the phenomenon, and focussing on a few, important aspects, the model can then help psychologists think about and explain the key processes involved in the phenomenon. This definition would apply to any kind of theoretical modelling, though models can often take a mathematical form, sometimes in the shape of a computer program. In this way, computers can be used as a tool in studying psychological processes – in particular, they are often used to study cognitive processes, such as perception.

Researchers can use models in several ways: firstly, they would use their existing knowledge to try to identify the essential features of, for example, perception or problem-solving. Initially, this might simply be put in the form of a 'flow diagram', representing the essential stages of the psychological process being studied, with arrows to show the flow of information etc. This theoretical model could then be taken a stage further through being represented mathematically, and perhaps even written as a computer program. The program could then be run to examine predicted performances, for any given input. Researchers then compare the performances of the program with human cognitive processes, to see how closely they match. Research using such models has advantages over research directly on the brain in that it is much simpler, and is also appropiate for detailed analysis (in a way we simply cannot do with the enormous complexity of the brain's network).

The need for precise specification in a computer program, arguably has the merit of forcing the cognitive scientist to explain clearly exactly how every aspect of their theoretical model actually operates. This has often shown just how sophisticated apparently 'simple' human activities are – such as perceiving a cup of tea and picking it up. Modelling these activities is a considerable challenge for cognitive scientists, and has been a valuable source of insight into some of the processes involved in such typical human activity.

The principles of psychological modelling have also been used to develop mathematical models that represent and simplify the complex interactions withina system of neurons in the brain (referred to as neural nets). Models of neural networks work on a quite different mathematical basis from cognitive models based on a precisely specifying every action taken. Models of neural networks use a more 'associative' type of model, where the patterns of interactions between different 'neurons' leads to an output which varies according to the stimulus. By being exposed to a range of stimuli, the model can go through stages of 'learning', which may shed some light on, for example, the ways in which human language is learnt.

Relationships

People