AQA Syllabus focus:
'The cognitive approach: the use of models to explain and make inferences about mental processes.'
Cognitive psychologists use models because thinking, memory, attention, and decision-making cannot be observed directly. Models provide simplified representations of hidden mental activity and help researchers connect behavior to explanations of cognition.
Why models are used in cognitive psychology
A central problem for cognitive psychology is that mental processes are private and internal. Psychologists cannot watch thoughts moving through the mind in the same direct way that they can observe a reflex or a visible action. Because of this, they use models to represent what might be happening between a stimulus and a response.
A model is a simplified representation of a mental process or system used to explain cognition and predict behavior.
Models are useful because they turn abstract ideas into something clearer and more systematic. They may show the sequence of processing, the parts of a system, or the flow of information from one stage to another. This makes it easier to test ideas scientifically.
Psychologists then compare the model’s predictions with actual behavior in experiments. If a model fits the evidence, it may be kept or refined. If it does not, it may be changed or rejected.
An inference is a conclusion about an internal mental process drawn from observable behavior or evidence.
In cognitive psychology, inferences are especially important because behavior is visible but cognition is not. For example, researchers may infer the existence of different processing stages from patterns of speed, accuracy, or error.
Types of models
Box-and-arrow models
A common cognitive model uses boxes and arrows.
Diagram of the Atkinson–Shiffrin (1968) multi-store model of memory, showing separate stores (sensory register, short-term store, long-term store) and the directional flow of information between them. It exemplifies how box-and-arrow models represent unobservable processing stages in a clear, testable sequence. Source
The boxes usually represent stages, stores, or components of processing, while the arrows show the direction of information flow.
These models often suggest that information moves through a system in an ordered way, such as:
input from the environment
processing of that information
output in the form of a response
Box-and-arrow models are popular because they are easy to understand and communicate. They help psychologists describe complex mental activity in a clear visual form. However, the boxes are not meant to be taken as literal structures in the brain. They are theoretical representations, not exact physical maps.
Computer models and the computer analogy
Cognitive psychology has often compared the human mind to a computer. Both can be described in terms of input, processing, storage, and output.
A labeled diagram of a basic computer system showing input, processing, output, and storage functions. This supports the computer analogy by making the shared information-processing vocabulary explicit and visually memorable. Source
This analogy helped psychologists think more precisely about how information might be coded, transformed, and retrieved.
Some researchers also create computer simulations of mental processes. If a computer program can perform a task in a way that matches human performance, this may support a particular explanation of cognition. For example, if the program shows the same kinds of errors or delays as people, psychologists may infer that similar processing principles are involved.
The value of the computer analogy is that it encourages scientific precision. The limitation is that people are not machines. Human thinking is also influenced by emotion, goals, meaning, and context.
How models explain and support inferences
Models help psychologists explain behavior by proposing what happens between stimulus and response. Instead of saying that a person simply reacts, a cognitive model suggests that information is attended to, interpreted, transformed, stored, or selected before a response occurs.
Researchers test these ideas using controlled methods. They may measure:
reaction time
accuracy
types of mistakes
changes in performance under different conditions
If performance changes in a reliable way, psychologists may infer something about the underlying process. For instance, a delay in responding might suggest that a task involves more than one stage of processing. A particular pattern of errors might suggest that information was encoded or retrieved in a specific way.
This means models do two related jobs:
they explain mental processing by offering a structure or sequence
they allow psychologists to make inferences from observable evidence
A good model should therefore do more than describe. It should also generate testable predictions. If those predictions are supported, confidence in the model increases. If not, the model must be reconsidered.
Strengths of using models
Using models has several important advantages in cognitive psychology.
They make the study of invisible mental activity more scientific and testable.
They provide a clear framework for organizing ideas about cognition.
They help researchers produce precise hypotheses and predictions.
They allow different findings to be linked into a single explanation of processing.
Models are especially valuable because they reduce complexity without abandoning explanation altogether. Without them, cognitive psychology would struggle to move beyond simple descriptions of behavior.
Limitations of using models
Despite their usefulness, models also have clear weaknesses. By definition, a model is a simplification, so it may leave out important features of real thinking. Human cognition is often more flexible and less orderly than a neat diagram suggests.
Some models can also become too mechanical, especially when based heavily on the computer analogy. Human beings are influenced by social meaning, emotion, motivation, and previous experience, and these factors may not fit easily into a simple information-processing system.
Another limitation is that the same observable behavior can sometimes support more than one model. Since inferences are indirect, psychologists cannot be completely certain that a model reflects the exact internal process involved. A model may fit the evidence well and still not be fully correct.
For this reason, cognitive psychologists treat models as useful tools, not perfect copies of the mind. Their value depends on how well they explain evidence, predict new findings, and remain open to revision.
Practice Questions
Outline what is meant by a model in cognitive psychology. (2 marks)
1 mark for stating that a model is a simplified representation of mental processes or a mental system.
1 mark for stating that it is used to explain cognition and/or make predictions or inferences about internal processing from behavior.
Discuss the use of models in cognitive psychology. (6 marks)
AO1 up to 3 marks:
Models are simplified representations of internal mental processes.
They are used because mental processes cannot be observed directly.
Cognitive psychologists use models such as box-and-arrow diagrams or computer-based models to explain information processing.
Models allow psychologists to make inferences from observable behavior such as reaction times, accuracy, and errors.
AO3 up to 3 marks:
Strength: models make cognition more scientific by generating testable predictions.
Strength: models provide a clear and organized way of explaining complex mental processes.
Limitation: models may oversimplify human cognition.
Limitation: the computer analogy can be too mechanical and may ignore factors such as emotion or context.
Limitation: different models may explain the same findings, so inferences may be uncertain.
FAQ
A verbal model explains a mental process in words, usually by describing stages or rules. It is useful when researchers want to define a theory clearly before testing it.
A diagrammatic model uses shapes and arrows to show parts of a system and how information flows. A computational model goes further by turning the theory into a computer program that can be run and tested against human performance.
Some models use separate stages because this makes it easier to isolate parts of a task and produce clear predictions. Stage models are often useful in early theory building.
Other models assume parallel processing because many real tasks involve several operations happening together. Researchers choose the format that best matches the evidence from experiments.
No. Brain imaging can support a model, but it does not prove the model is fully correct.
A scan may show that certain tasks activate different areas, which can fit a model’s claims about separate processes. However, a model is still judged mainly by how well it explains behavior and predicts results across studies.
Observable behavior is only the final outcome of hidden processing. Different internal systems can sometimes produce the same response pattern.
This is why psychologists design new experiments to separate the models. They look for conditions where one model predicts a different result from the other, such as a different error pattern or timing effect.
A model is usually revised when it still explains a lot of evidence but needs adjustment to handle new findings. Small changes may improve its accuracy without abandoning its core ideas.
A model is more likely to be replaced when it repeatedly fails to predict results, becomes too complicated to defend, or is outperformed by a simpler and more powerful alternative.
