Flashcards
Sem título
Flashcards do conjunto (46)
- 1Define semantic memory.Conceptual knowledge
Aspect of memory corresponding with knowledge of objects, word meanings, facts and people, without connection to time or place. - 2What is a common breakdown of semantic memory?Dementia
- 3What did early studies on semantic memory organisation use in their experiments?Sentence verification task:
DV = RT
Need to identify a statement as T/F
Two sentence types:
- Set inclusion: subject belongs to category
- Property-attribute: describes properties of a subject - 4What characterises network models?Concepts represented by nodes, with relationships between them represented by links
Set inclusion = 'is a' link
Property attribution = 'has' link - 5Explain the Hierarchical Model (Collins & Quillian, 1969).Concepts organised hierarchically
Idea of a 'cognitive economy' - property attribution stored in the most general level, not necessarily with the exact concept --> requires use of inference
Thus, sentence verification is a function of levels --> RT should increase with complexity - 6What problems did Conrad (1972) find with the hierarchical model?RT better explained in terms of frequency of co-occurrence of concept + property, rather than levels i.e. salient features
Within-category typicality effect
Faster negative judgements for closer (more similar) concepts - 7How does the Spreading Activation Model (Collins & Loftus, 1975) differ from the hierarchical model?Concepts not organised hierarchically
Links vary in associative strength and accessibility, which explains the typicality effect.
Activation of a concept spreads to other linked concepts (would explain the semantic priming effect). - 8What is the typicality effect???
- 9What is the semantic priming effect (Myer & Schvaneveldt, 1976)?RT for a word faster following a semantically related word (prime) e.g. car-truck vs. flower-truck
Used to study organisation of semantic memory and automaticity - 10Describe the Feature Comparison Model (Smith, Shoben & Rips, 1974).Network models assume localist representation but feature comparison assumes knowledge is represented as distributed features in semantic space.
Defining vs. Characteristic features
Two-stage decision process:
1. Compare all features of subject and predicate: high/low similarity = T/F, intermediate similarity -->
2. Compare Defining features for match/mismatch - 11How can we test the Feature Comparison Model?Multidimensional scaling: subjects rate how similar a pair of subjects are
Linguistic hedges: true or false
Problems with clear definition of defining/characteristic features - 12What are some strengths and weaknesses of sentence verification tests?Important data in developing theories of semantic memory, esp. typicality and similarity effects
Limitations in assessing structure and process
Can be explained by various assumptions - 13What two main types of neuropsychological studies have been conducted?1. Selective impairment of categories (living vs. non-living) --> more common to have living things impaired, with non-living things intact
2. Category specific impairment e.g. preserved knowledge of body parts + non-living things - 14What did Warrington and Shalice's (1984) case study of Patient JBR demonstrate?Densely amnesic with intact language and speech
Consistent pattern of impairment across a variety of tasks to tap semantic memory e.g. naming pictures, giving a definition
- Impaired knowledge of living things + preserved knowledge of non-living things
More specific impairment of musical instruments and food category - 15How does Perceptual Functional Theory (Warrington & Shalice, 1984) describe the organisation of concepts in the brain?Category-specific impairments arise due to different types of distinguishing properties between category members
- Living things: perceptual (visual) features
- Non-living things: functional basis
Problems: patients not necessarily impaired for one type of knowledge; some properties are neither - 16Why is Distributed Plus Hub Theory preferable?Hub for each concept/object in addition to distributed modality-specific info.
This hub is 'amodal' - not tied to functionality or perception
Provides an efficient way of integrating our knowledge of any given concept
Makes it easier to detect semantic similarities across concepts differing greatly in their modality-specific attributes - 17What do impairments in colour and object recognition demonstrate about the distributed-plus-hub theory?E.g. SD patients chose 'green pumpkin' - green shared by a no. of vegetables (that knowledge is distributed)
Not recognising pumpkins as orange (idiosyncratic) shows a lack of the hub
Delayed-copy drawing: lack specific knowledge e.g. camels have humps - 18What are categories and how do we utilise them?Class of stimuli that are treated in an equivalent manner
Serves to cut down the diversity of objects and events that we deal with differently in an organism of limited capacity - 19What are natural categories?Categories that occur in natural language
Correlation structure i.e. features not combined arbitrarily, as in artificial categories - 20What are the key characteristics of natural categories?Boundaries ill-defined i.e. some items will have low consistency e.g. pumpkin-vegetable
Family resemblance: no attributes shared by all members of the category but each member has at least one attribute in common
Typicality: attributes of a 'typical' member of that family (high correlation with family resemblance score)
Internal structure: typicality gradient i.e. some members are better than others, with the category centred around a 'prototype' (most typical member) - 21What are basic level categories?Most inclusive level
Most informative and most frequently used e.g. chair
Superordinate = more general e.g. furniture
Subordinate = more specific e.g. office chair - 22What is the prototype view?New instances compared to the prototypes
- 23What is the exemplar view?Unlikely to have prototypes for more abstract concepts so have to compare individual instances
Exemplar and prototype views makes similar predictions for natural categories - 24Theory (Knowledge) Based ViewSuggests categorisation is based on a theory developed from specific instances
Context-dependent e.g. circular object with a diameter of 6cm seen as more likely to be a pizza - 25Why is reading not 'natural'?Culturally acquired skill
Needs formal instruction, unlike speech - 26What are some research methods involved in studying reading?RT tasks - lexical decision, semantic categorisation, naming
Indirect measures: interference tasks, priming
Record eye movements during reading
Neuropsychology
Neuroimaging - 27Is word identification automatic? How do we know?Fairly.
Stroop task
Masked priming: prime presented briefly and masked so that participants cannot identify it but still affects their response - 28What are the different levels of linguistic processing?Orthography (spelling)
Phonology
Semantics
Syntax
Higher level discourse integration (making inferences) - 29What is McClelland and Rumelhart's Interactive Activation Model (1981)?Hierarchical:
- Feature level: feature detected, activation is sent to all letters containing that feature e.g. vertical line
- Letter level: once a letter is identified, activation to all words containing it in the appropriate position (other words inhibited)
- Word level: activated word units increase level of activation in the letter-level units for the letters forming that word (recognised at a particular threshold) - 30Explain slot coding.Assumption that wrongly predicts that letters with correct first and last letter can still be processed.
Cannot capture similarity between words with letters in wrong order e.g. CAT not = to ACT - 31What is the transposed letter priming effect?In masked priming task, response to target (e.g. Judge) is faster if primed by a transposed letter prime (e.g. Jugde) than by a substituted letter prime (e.g. Junpe).
- 32What is Frost's (1998) strong phonology hypothesis?Phonological representation is a necessary product of processing printed words
- 33What is the homophone interference effect (Van Orden, 1987)?Subjects asked to categorise homophones by their semantics - more errors when asked to make a semantic categorisation e.g. ROWS mistaken for ROSE
- 34What is masked phonological priming?Word processing faster when preceded by phonologically identical non-word primes than by primes similar in orthography but not phonology (shows we process words phonologically)
- 35What is orthographic depth?How predictable spelling-to-sound mappings are
- Deep/opaque: inconsistent, ambigious e.g. tough, though, cough
- Irregular mapping e.g. pint vs. mint
- Shallow/transparent: one-to-one mapping - 36What do we need in order to pronounce an irregular word correctly?Stored pronunciation where spelling to sound mapping words is not possible
- 37What are the key assumptions of the Dual Cascaded Model (Coltheart et al., 2001)?Individuals use both the non-lexical and lexical paths when reading aloud.
Non-lexical route insufficient for irregular words
Naming visually-presented words primarily depends upon lexical route (faster)
Activation at one level cascades to the next before processing at the first level is complete - 38What are the different types of dyslexia?1. Acquired dyslexia: due to brain damage
- Surface: intact non-word reading but poor at irregular words
- Phonological:
2. Developmental dyslexia: individuals who have always struggled - 39How does the study of individuals with acquired dyslexia demonstrate the assumptions of the dual-cascaded model?Route 1: Grapheme-Phoneme Conversion - Allows accuracy for non-words even if semantic knowledge is impaired (surface dyslexics)
Route 2: Lexicon & Semantic System - phonological dyslexics impaired in Route 1, so need sight vocab. in their memory -> struggle with unfamiliar words
Route 3: Semantic system is bypassed - 40What are some concepts within eye movement studies?Saccades: jump from one feature to another (usually approx. 7-8 characters)
Fixations: 250ms following a saccade
Regressions: infrequent right to left movements of approx. 2-5 characters
End of line sweep: long single saccade from characters at the end of one line to the next - 41How large is our perceptual span?Effective field of view: 3-4 letters to the left of fixation and up to 15 letters to the right
- 42Which words are most likely to be skipped?Common, short or predictable
20% of function words: articles, conjunctions, pronouns - 43Describe the reading rate in normal reading.Differs across different types of material
Readers control reading rate within a sentence as a function of comprehension difficulty
Fixation longer for more difficult words and low frequency words
Shorter saccades, longer regression - 44What are Grice's (1967) Components of Effective Communication?Cooperative principle
- Maxim of quantity: informative as is necessary
- Quality: truthfulness
- Relation: relevance
- Manner: easy to understand - 45What errors reveal the regularities of speech (Meringer)?Anticipations e.g. Take my bike --> bake my bike
Perservations e.g. he pulled a tantrum --> pantrum
Exchanges e.g. getting your nose remodelled --> model renosed
Substitutions - 46Describe the Two Stage Model of processing.Conceptual Semantics --> (Lexical Concept) --> Lexical Selection --> (Lemma) --> Morphological encoding --> (Morpheme) --> Phonological Encoding --> (Phonological Word) --> Phonetic Encoding --> Articulation
Informações
- Descrição
- FlashCards sobre Sem título, criado por ashleigh.d1 em 14-06-2015.
- Criado por
- ashleigh.d1
- Visualizações
- 0
- Avaliação