L12 - Cognitive Models

Question 1

Generally speaking, what is a model?

Answer 1

Theories or potential explanations.

Question 2

What is the focus of cognitive models?

Answer 2

To propose potential explanations for the cognitive processes that are taking place within our brain that enable us to make sense of the world.

Question 3

What is the information processing approach to models?

Answer 3

Process models indicate the flow of information through a system (like a computer).

Boxes are generally used to indicate the different processes, and arrows indicate directional flow.

A more informal type of model​ (E.g. 3 stage model for memory)

Question 4

Why is the information processing approach for making models useful?

Answer 4

1. Useful as a means for describing and understanding the different processes associated with a given phenomenon (memory, attention etc.)
2. They allow researchers to form specific hypotheses and test those hypotheses.

Question 5

What are the names of these two models and who were they made by?

Answer 5

1. Broadbent’s (1959) Early Selection Model of Attention
2. Treisman’s (1964) Attenuator Model of Attention

Question 6

What is the mathematical approach to models of the cognitive process?

Answer 6

When researchers formally state hypotheses in a mathematical sense.

Question 7

What is superior for hypothesis testing, information processing approach or mathematical approach?

Why?

Answer 7

Mathematical approach.

It relies on empirical data (behavioural responses of participants in an experiment) and a quantitative approach to assessing competing hypotheses.

Question 8

Why was the advent of the computer and computer simulations useful for psychological researchers? (3 reasons)

Answer 8

1. They provided a useful framework (metaphor) for thinking about cognition
2. They provided the opportunity to formally define and implement computation models of cognitive processes.
3. Computer simulations force researchers to change from vague terminology (e.g. ‘fusion’, ‘stimulus features’) to precise terminology that could be put into a computer.

Question 9

Describe a Category Learning Task.

Answer 9

1. Participants are shown examples of stimuli from 2 categories
2. They are asked if a stimulus is from A or B
3. They are given feedback, and they learn the categories over time
4. They are then shown new stimuli drawn from the same category structures and asked to classify them.

Question 10

Describe The Generalized Context Model; GCM (Nosofsky, 1986) that is used to predict ‘category learning task’ results.

Answer 10

The probability of a stimulus being categorized as a member of a given category is a weighted function of the distance between the target stimulus and the members of the two categories in the space.

Question 11

For two-choice decision making for computational models, what theory does a good job of describing decision making data?

What does it do?

Answer 11

Signal Detection Theory

It accumulates evidence in favour of one decision or the other until it reaches a decision boundary

Question 12

What makes a model a ‘good cognitive model’?

Answer 12

1. Understanding that a model is an exploration and not the real thing
2. A given model means nothing in and of itself (we can only judge models in relation to other models)
3. It must have some sort of evaluation criteria.

Question 13

What did Jacobs and Grainger (1994) suggest that models should be judged in terms of? (4 things)

Answer 13

1. Descriptive adequacy (the model should provide a good fit to empirical data)
2. Complexity (Simple models are good models)
3. Generality (good models should generalize across multiple experimental settings and manipulations)
4. Explanatory adequacy (how plausible is the model?)