Flashcards in Knowledge Deck (44):
What is a concept?
A concept is a mental representation that is used for a variety of cognitive functions, including memory, reasoning, and using and understanding language (Solomon et al., 1999).
You think of a cat. Use the idea of concepts to explain what is happening.
When you think about cats, you are drawing on your concept, or mental representation, of cats, which includes information about what cats are, what they usually look like, how they behave, and so on.
Concepts are interesting to cognitive scientists. What type of concepts are most thoroughly studied in cognitive psychology?
Categorization, which is the process by which things are placed into groups called categories.
What is one of the most important functions of categories?
That they help us to understand individual cases we have never seen before. For example, being able to say that the furry animal across the street is a "cat" provides a great deal of information about it.
Categories have sometimes been called ...
"pointers to knowledge" (Yamauchi & Markman, 2000). Once you know something is in a category, you know a lot of general things about it.
How have cognitive psychologists thought about how objects are placed into categories?
1. We use a definitional approach.
2. We use a protoype approach.
3. We use an exemplar approach.
What is a prototype approach to object categorisation?
According to the prototype approach to categorisation, membership in a category is determined by comparing the object to a prototype that represents the category.
What is a exemplar approach to object categorisation, and how does it relate to the prototype approach?
The exemplar approach to categorisation, like the prototype approach, involves determining whether an object is similar to a standard object. However, whereas the standard for the prototype approach is a single "average" member of the category, the standard for the exemplar approach involves many examples, each one called an exemplar.
What is a definitional approach to object categorisation?
According to the definitional approach to categorisation, we can decide whether something is a member of a category by determining whether a particular object meets the definitions of the category.
There is a problem with the definition approach to categorisation. What?
Definitions are usually too narrow or not narrow enough to account for the intuitive categorisation people do.
So.. What is a "typical" member of a particular category?
Elinor Rosch (1973) proposed that the "typical" prototype is based on an average of members of a category that are commonly experienced.
Researchers have looked at how typical a member of a particular category is. What predicts how typical an item is for a category?
Rosch and Mervis (1975) had participants write down common characteristics and attributes for different objects of the same category (e.g. furniture). Those that had a lot of similar attributes/characteristics were more likely to be regarded as typical for that group.
What are the research findings we have from research on prototypes?
1. Prototypical objects have high family resemblance.
2. Statements about prototypical objects are verified rapidly.
3. Prototypical objects are named first.
4. Prototypical objects are affected more by priming.
Statements about prototypical objects are verified rapidly. Huh?
Edward Smith and coworkers (1974) used a procedure called the sentence verification technique to determine how rapidly people could answer questions about an object's category (e.g. "An apple is a fruit" - yes/no).
Participants reponded faster for objects that are high in prototypicality than they did for objects that are low in prototypicality.
Prototypical objects are named first. Huh?
When participants are asked to list as many objects in a category as possible, they tend to list the most prototypical members of the category first.
How are prototypical objects affected more by priming?
Rosch, 1975b gave participants a primed color (e.g. "green"), then would show them slides of either a prototypical green or a less prototypical green, or a different color entirely. They were told to indicate whether the primed color was same to the color on the slide. Reponse time was faster on prototypical colors.
What approach works better: prototype or exemplars?
Based on the results of a number of research studies, some researchers have concluded that people may use both approaches. It has been proposed that as we initially learn about a category, we may average exemplars into a prototype; then, later in learning, some of the exemplar information becomes stronger. Thus, early in learning we would be poor at taking into account "exceptions", such as ostriches or penguins (birds), but later, exemplars for these cases would be added to the category.
Explain Rosch's research on category levels.
She distinguished three levels of categories, and made an experiment in which she had people list characteristics and features of an object in the different level categories (furniture -> chairs -> kitchen chairs). She found that there was a discrepancy in the amount of information people listed for the different levels, where the specific level had the most information, but the basic level had almost as much. She also found that when shown pictures of specific objects and being asked to say a word that identifies each picture, they would pick words from the basic level.
What are the different category levels of Rosch?
1. Global level (furniture)
2. Basic level (chairs)
3. Specific level (kitchen chairs)
How do knowledge affect categorisation?
Knowledge affects categorization in many ways. According to a prototype approach to categorization, more knowledge about the category would probably change the prototype as it is an average of all items in that category. With the exemplar approach it probably would also matter, because more exemplars would be available with more knowledge of the category. Some researchers believe that we rely more of prototypes when we have less knowledge, and more on exemplars when we have much knowledge. Rosch's level of categorization, research has showed that experts are more likely to use a word from the specific level on items that they are knowledgeable in. This differs from the general trend of using the basic category.
What is Collins and Qullian's hierarchical model?
It is one of the first semantic network models to be developed. (1967, 1969). The network consists of nodes that are connected by links. Each node represents a category or concept, and concepts are placed int he network so that related concepts are connected. In addition, properties associated with each concept are indicated at the nodes.
What predictions could be made from Collins and Quillian's hierarchical model?
Because of the network's hierarchical organisation, in which general concepts are at the top and specific ones are at the bottom, is that it results in the testable prediction that the time it takes for a person to retrieve information about a concept should be determined by the distance that must be traveled through the network.
Because of Collins and Quillian's network's hierarchical organisation, in which general concepts are at the top and specific ones are at the bottom, is that it results in the testable prediction that the time it takes for a person to retrieve information about a concept should be determined by the distance that must be traveled through the network. Did this hold true?
How can Collins and Quillian's hierachical model account for some effects of priming?
According to the idea of spreading activation, activation of a link between nodes leads to the activation spreading further than the target movement in the network. This causes nodes in close proximity to the end node to become primed and can so be retrieved more easily from memory.
According to the idea of spreading activation, activation of a link between nodes leads to the activation spreading further than the target movement in the network. This causes nodes in close proximity to the end node to become primed and can so be retrieved more easily from memory. Has this been tested?
Yes, naturally. Myer and Schvaneveldt (1971) used a variation of the lexical decision task by presenting participants with two strings of letters, one above the other. The task was to press, as quickly as possible, the "yes" key when both strings were words or the "no" key when one or both were not words. The categories of the words should not matter, but results showed that participants were faster to press yes when both words had categorical similarities.
Mention a phenomenon of categories that is not accounted for by the Collins and Quillian model.
We know that the typicality effect, in which reaction times for statements about an object are faster for more typical members of a category than for less typical members. This is not explained by the model because the travel time is equal between the nodes.
What happened to Collins and Quillian's model?
It got revised by Collins and Elizabeth Loftus into their own semantic network which focuses on how personal experience affects networks.
Outline the Collins and Loftus semantic network model.
The Collins and Loftus model abandons the hiarachical structure used in Collins and Quillin' in favor of a structure based on a person's experience. Concepts that are more closely related are connected by shorter lines. These shorter links predict faster reaction times for more typical objects. The spacing between various concepts can differ for various people depending on their experience and knowledge about specific concepts.
What is the main critisism of Collins and Loftus' model?
That is is so flexible that it becomes difficult to falsify. The idea that the length of the links predicts how fast nearby concepts are activated is worthless when the model designs the links to be of different lengths for different people.
For your exam, you need to know three approaches to explaining how knowledge is represented in the mind. Which?
1. Collins and Quillian's Hierarchical model.
2. Collins and Loftus Model
3. The connectionist model
James McClelland and David Rumelhart proposed a new approach to explain how knowledge is represented in the mind called connectionism. Why did this approach gain favor among researchers?
1. It is based on how information is represented in the brain.
2. It can explain a number of findings.
What is connectionism?
Connectionism is an approach to creating computer models for representing concepts and their properties based on characteristics of the brain.
Connectionist models are also called...
Parallel distributed processing (PDP).
What are the units of a connectionist model?
They are based off how neurons act.
What is the rough outline of a connectionist network?
It features input units that activate hidden units, that in turn active output units. The connections between units are weighted to either increase or decrease the activity of the target unit.
In a connectionist network, what determines the activation of a unit?
1. The signal that originates in the input unit.
2. The connection weights between units.
In a connectionist model of the mind, how is any given stimulus represented?
A stimulus presented to the input units is represented by the pattern of activity that is distributed across the other units. This pattern may be unique for each stimulus.
What is back propagation in connectionist models?
Back propagation is the process by which error signals are sent back to the hidden and representation units to provide information about how the activation weights should be changed so that the correct property units will be activated.
What are the strengths of a connectionist model?
1. The operation of connectionist networks is not totally disrupted by damage.
2. Connectionist networks can explain generalization of learning.
The operation of connectionist networks is not totally disrupted by damage. Explain.
Because information in the network is distributed across many units, damage to the system does not completely disrupt its operation.
Connectionist networks can explain generalization of learning.
Because similar concepts have similar patterns, training a system to recognise the properties of one concept also provides information about other, related concepts.
What is the greatest critisism of the connectionist approach?
That the connectionist models do not agree on how a back propagation error-correcting mechanic looks like. It has also been difficult to find a similar mechanism in the brain.
Different categories of objects are associated with activity in specific areas of the brain. Which seems to respond strongly to faces and where is it located?
The fusiform face area (FFA). The FFA is located in the ventral stream on the ventral surface of the temporal lobe on the lateral side of the fusiform gyrus.