Cognitive Pysch Exam 3

knowledge

explicit awareness obtained and mentally available info about the world and ourselves that was obtained through experience

conceptual knowledge

knowledge that enables us to recognize objects and events, and make inferences about their properties

conceptual knowledge allows us to

interact and know/assume

in class example about conceptual knowledge

we have an idea of a sandwich and a hotdog doesn’t fit the definition

concepts

mental representations used in cognitive functions

categories

meaningful grouping

all possible examples of a particular concept

categories (2nd def)

categorization

process of forming categories

forming categories

definitional approach, family resemblance, prototype approach, exemplar approach, knowledge based view, hierarchial organization

forming categories: definitional approach

does it meet the definition of the category?

what does the definitional approach work for

shapes

what does the definitional approach not work well for

man made objects

forming categories: family resemblance

members resemble each other, have characteristic features

what does family resemblance allow for

variation

forming categories: prototype approach

for every category there is a prototype which is the best example

prototype

typical member of category

high prototypicality

closely resembling prototype

low prototypicality

does not resemble protoype

in class example of prototype approach

ranking birds

what type of prototypes are verified rapidly

high proto

which types of prototypes have many characteristic features

features: high proto

sentence verification the TYPICALITY EFFECT

high proto has a fast repsonse and low proto is a slow response for sentence verification

which categories are named first and what is it effected by

high proto, effected by priming

forming categories: exemplar approach

you store exemplars from the past (like a photo album) and pull them for reference

exemplar

members encountered in the past

problem with exemplar

our categories vary (example: games)

prototype versus exemplar

we use both!

prototypes in exemplar approach

quick decision, learning categories

exemplars in exemplar approach

atypical cases, variable categories

forming categories: knowledge based view

use knowledge to justofy classficiation

what does knowledge based view use

perosnal and schemas

forming categories: hierarchical organization

overarching category (like an animal) with other categories underneath (like a bird)

general flow of hierarchical organization

starts off broad and gets more specific

general level in hierarchial organization is…

the global level

order of levels in hierarchial organization are…

general → basic → specific

what type of features are named at a basic level

common features

naming task shows

we name at a basic level and that basic level is special

what is faster if given a basic level

categorization

influence of knowledge in knowledge based view

experts use the specific level instead of the basic

organization: semantic network approach

concepts arranged in networks, interconnected by category

Semantic network approach: Collins and Quillian Model

nodes, links, is hierarchical goes from broad to specific,

Collins and Quillian Model: what are nodes

represent categories and concepts that contain associated features

:Collins and Quillian Model: what are links

indicate relations between nodes

Collins and Quillian Model: cognitive economy

shared features are stored at higher level nodes and specifics are stored at lower level nodes

Collins and Quillian Model: do nodes and links repeat?

no, there is no reduncy

Collins and Quillian Model: model prediction

RT should reflect distance in model (how many levels you travel)

Collins and Quillian Model: small RT

close nodes produce this

Collins and Quillian Model: how do you test the model prediction

sentence verification task (example: a canary is a canary should be the fastest)

Collins and Quillian Model: property of the model

spreading activation

Collins and Quillian Model: what is activation

the selected concept

Collins and Quillian Model: what is spreading activation

any link connected to selected node also becomes active

Collins and Quillian Model: what is a result of activation

result is the priming effect

CogLab 11 Lexical Decision: task

is this a word yes/no

CogLab 11 Lexical Decision: DV

measuring RT

CogLab 11 Lexical Decision: IV

word, nonword, string of related or nonrelated words

CogLab 11 Lexical Decision: results

faster RT for related words, shows how spreading activation is due to priming effect

problems with Collins and Quillian

cannot predict typicality effect (difference in RT due to prototypically)

Organization: connectionist model

computational model for representing concepts and properties, inspired by brains neural network

how is the connectivist model like a brain

interconnected units (neurons), units do simple tasks (send signals), work in parallel and are distributed

connectionist model: output unit

contain info, result of computation

connectionist model: hidden unit

activated when selected, computation

connectionist model: input unit

can be activated by stimuli or by other units

connectionist model: links/connections

connect units, transfer info between, like axons

connectionist model: connection weights

each line is weighted -1 to 1

connectionist model: green connection line

1, is the excitatory connections (turns on)

connectionist model: red connection line

-1, inhibitory connection (turns off)

connectionist model: black connection line

0, no transfer (neutral)

connectionist model: how is stimulus represented

pattern of activity distributed across units

connectionist model: how does model learn

back propagation

connectionist model: backpropagation

process of changing connection weights in repsonse to error, repeats unit model until it is correct

connectionist model: advantages

can replicate the mind by stimulating areas of coginition, generalize learning, explain typicality effect, not disrupted by damage

connectionist model: graceful degradation

performance decreases gradually as system is damaged

connectionist model: disadvantages

propagation can be too powerful, not exactly like a brain, doesn’t make mistakes like humans

what is the best model of cognition so far?

connectionist model

RR Chat GPT: overview

compared AI against humans in test-retest and rating scales for typicality, found AI is moderatly similar to humans but not the same

RR Chat GPT: what type of model is AI

it is a connectionst model

RR Chat GPT: does AI represent categoriess like humans

may not represent the same, but the output is similar

RR Chat GPT: how can Ai effect data collection

fabrication and not human responses

language

communication system taht uses symbols to express feeling, thoughts, ideas, and experiences

examples of types of languages

spoken, written, and signed

psycholinguistic

study of language and how humans process

comprehension of language

understanding

production of language

producing speech

representation of language

how you show (like writing)

acquistion of language

how you learn and develop it

langauges throughout cultures common qualities

develop in same stages, same foundation, creative, transmission, hierarchical, and rules

transmission of language

passed down through generations, changes over time

example of language being transmissible

slang

language’s hierchical structure

smalle runits combine to make bigge units

rules in language

determine how units can combine, we are often unaware of rules

phonems

sound, smallest unit

example of phonem

/k ae t/

rules for phonems

known but not aware (we know tn isnt a letter combo)

words

stored in mental lexicon

examples of what is stored in the mental lexicon

dictionary, terms, pronunciation, and meaning

syntax

rules on how to combine words

identifying words

knowledge of patterns, meanings, context and frequency

what helps in idenifying words

context and frequency (top down processing)

top down processing in idenitfying words

expectations matterm and so do prior lexical knowledge

example of top down processing in language

lauren versus yanny (hearing one becasue you know the trend)