Cognitive Psychology Lecture Notes PDF

Knowledge In today’s lectures, we will discuss: Explain concepts and categories Three ways of categorisation...

Knowledge In today’s lectures, we will discuss: Explain concepts and categories Three ways of categorisation  The Definitional Approach  The prototype approach  The exemplar approach KMF 1023 Relationship between Categories COGNITIVE PSYCHOLOGY  Semantic Networks Lecturer: Norehan Zulkiply FSKPM How are categories stored in the brain Universiti Malaysia Sarawak Knowledge Organisation through Knowledge Organisation through Concepts & Categories Concepts & Categories Concepts and Categories Basic important processes that allows us to car bicycle lettuce organise our experience of the world so that we can function in it carrots What are concepts? What are categories? truck broccoli train peas bus corn Knowledge Organisation through Knowledge Organisation through Concepts & Categories Concepts & Categories Item Vehicle Vegetable Concepts – mental representation 1 car peas consisting of general idea or understanding 2 truck carrots Themeaning of words and symbols 3 bus broccoli facts about the world 4 train lettuce What objects look like 5 bicycle corn Stored in Long term memory 1 Knowledge Organisation through Knowledge Organisation through Concepts & Categories Concepts & Categories How do you know you have understood a Categories – placing things into groups, or a class of similar things concept? Categories have been called ‘pointers to If a person has mastered a concept, knowledge’ he/she should be able to categorize  provide basic knowledge about concepts objects or events of a domain  valuable tool for making inferences about members from the same category  help understand new situations How do we Categorise? Three possible ways: Three Ways of Categorising Concepts 1. through definitions 2. through prototypes 3. through exemplars 1. The Definitional Approach 1. The Definitional Approach Based on determining whether the E.g.: “Cat” E.g.: “bachelor” properties of a particular object match a  Has fur  “adult”, definition  Likes milk, fish  “human”,  Dislikes water Categorisation through definitions  “male”  Has four legs E.g....  “unmarried”  Meows  Has a tail  Has whiskers  Chases mice  Belongs to the cat family 2 1. The Definitional Approach Wittgenstein (1953)  E.g. “chair”  a piece of furniture Proposed family resemblance – to deal  consisting of a seat, legs, back, often with the fact that definitions often do not arms, include all members of a category  designed to accommodate one person BUT  Things in a particular category resemble one  beanbag & car seat do not meet our another in a number of ways. definition Limitation of using the definition Instead of setting definite criteria that approach: Definitions do not, and every member of a category must meet, cannot, include all members of a the approach allow for some variations category within a category 2. The Prototype Approach Bird Example Eleanor Rosch (1973) We decide what belongs to a category by determining the similarity between the object in question and a prototype Prototype: a “typical” member of a particular category; the average representation of the category (formed from averaging the category members After having seen lots of birds in your life you would have formed a prototype of what a typical encountered in the past) bird would look like Prototypicality Eleanor Rosch (1973) incorporated the Variations within categories represent Wiitsgentein’s idea of family resemblance differences of prototypicality into the prototype approach to the study High-prototypicality - category member of categorization closely resembles the category prototype  allow variations within a category Low-prototypicality - category member distantly resembles the category prototype 3 Quantified Prototypicalities (Rosch, Quantified Prototypicalities 1975) Typicality ratings for members of three categories (Rosch, 1975) Quantified this idea by presenting a category title, such as furniture, and a list of 50 members of the category Participants were ask to rate the extent to which each member represented the category title on a 7 point-scale. Rating 1-means that member is a very good example of what that category is Rating 7 – means member fits poorly within the category/ is not a member at all Effects related to prototypicality Prototypical objects have high family resemblance They have more in common with other members of the category Prototypical objects have high family Rosch & Mervis (1975) – demonstration resemblance - ask participants to list as many attributes that they feel are common to the following objects: Statements about prototypical objects are  chair verified rapidly (reaction time faster in high  sofa  mirror protypical items than to low)  Telephone Results: Prototypical objects are named first  Characteristics overlap: many for chair and sofa ( means that family resemblance of these items is high), far less for mirror and telephone (family resemblance is low)  Chair, sofa –indicate good examples of the furniture (share many attributes with other members of the category). While mirror and telephone are bad examples of category furniture Statements about prototypical objects Sentence-verification Data: are verified rapidly Sentence-verification technique (Smith et al., 1974) Answer yes if you think the statement is true:  An apple is a fruit A pomegranate is a fruit Participants responded faster for objects that are high in prototypicality (first statement) 4 Prototypical objects are named first 3. The Exemplar Approach When participants are asked to list as Involves comparing object seen in a new many objects as possible, they tend to list instance to examples (whether the object is the most prototypical members of the similar to a standard object) category first (Mervis et. Al., 1976) Exemplars: examples of actual members of the Name some vehicles: category, previously encountered  car  truck This approach assumes that people categorize  plane new instances by comparing them to  train representations of previously stored examples in  submarine memory 3. The Exemplar Approach Which Approach Works Best? Eg. You have already seen Both prototype and exemplar approaches have their advantages and disadvantages Some researchers believe that both approaches are used Others suggest exemplars work best for small categories and prototypes work best So how would you categorize this? for large categories New Instance Categories – is there a privileged level? Rosch’s Approach (1976) Hierarchical categorisation Basic level is ‘special’ – above this level Which level is more important in this much information is lost, and below this hierarchy? Is there a privileged tier? level little information is gained 5 Demo of Rosch’s Approach How Knowledge Affects Categorisation e.g naming things For the following categories, list as many features that would be common to all, or most, of Stimuli for the “naming things” demonstration the objects in the category. For example, for E.g. Named guitar (basic level) rather than table you might list ‘legs’ musical instrument (superordinate) or rock guitar Results: (subordinate) How Knowledge Affects Categorisation How Culture Affects Categorisation Experts vs. Nonexperts Maya Itza, in Guatemala  identify plants and animals at the subordinate level (due to close contact with natural environment)  E.g an oak tree is classified as an oak tree rather than a tree Results of J. W. Tanaka and Taylor’s (1981) “expert” experiment. Experts (left pair of bars) used more subordinate categories to name birds and nonexperts (right pair of bars) used more basic categories. Relationships between Categories How are categories organized? Semantic networks : Relationship between Categories  concepts arranged in networks  arrangement represents organization in the mind  Based on Collins & Quillian’s (1969) research  Model of memory suitable for computer simulation 6  Components of Semantic Networks Components of Semantic Networks Nodes Links Concepts Properties of a concept Building a semantic network. (a) The skeleton-nodes connected by links. (b) Adding concept names to the nodes, which more specific ones at the bottom and more general ones at the top. Components of Semantic Networks Cognitive Economy in semantic network Store shared properties at a higher- level node  Exceptions added at lower nodes when necessary (c) Adding properties of each concept. This is the Why? Storing e.g “can network proposed by fly” at the node of Collins and Quillian every bird (canary, (1969). robin, etc)is inefficient and would use up too much storage space Testing Collin’s & Quillian’s Model Testing Collin & Quillian’s Model Provides testable predictions for concept recall Results: Greater distances are associated with longer reaction times, both when verifying statements about The time to recall information is determined by properties of canaries (top) and about categories of which the distance between info in the network canary is a member (bottom) Example...”A canary is an animal”, “A canary is a bird” 7  Theory of Spreading Activation Property of Collin’s & Quillian’s model Spreading activation -- activity that spreads out along any link that is connected to an activation node How are categories stored in the brain Categories in the Brain Categories in the Brain Specific or distributed? Areas of the brain are selective for categories (based on Different categories of objects are represented neuropsychological research) by activity in the specific areas of the brain Warrington & Shallice (1984):  Examples of areas for categories inferior temporal (IT) lobe damage related with inability to FFA –responds to faces recognize living things while retaining the ability to PPA – responds to houses, rooms, places recognize human made artifacts (such as tools and E.g., the representation of a cat activate: furniture)  sensory areas (for how a cat looks like) later research find some patients who had difficulty  Motor area ( for how it moves) recognizing tools, but not living things  Higher level areas (for knowledge about the cats’  Visual Agnosia –can see objects perfectly well, but they cannot name these objects behavior)  Emotional areas (for the emotional response elicited by the cat) Summary – What do you need to know? Summary Differences between concepts and Relationships between Categories categories  Collins & Quillan’s Semantic Networks Theory of spreading activation Three ways of categorisation  The definitional Approach How are categories stored in the brain  The prototype approach  The role of the IT lobe  The exemplar approach How are categories affected by  having expert knowledge  Different cultures 8