Cognitive Psych Notes Week of 3/30
Cognitive Psych Notes Week of 3/30 PSY 0422
Popular in Cognitive Psychology
Popular in Psychlogy
This 6 page Class Notes was uploaded by Lindsay Notetaker on Friday April 3, 2015. The Class Notes belongs to PSY 0422 at University of Pittsburgh taught by in Spring2015. Since its upload, it has received 160 views. For similar materials see Cognitive Psychology in Psychlogy at University of Pittsburgh.
Reviews for Cognitive Psych Notes Week of 3/30
Report this Material
What is Karma?
Karma is the currency of StudySoup.
Date Created: 04/03/15
330 Concepts are organized by theories To decide if something is a member 0 Not match of features rather 0 Ask if it ts Role of similarity o A good heuristic guess 0 Deep and surface are correlated 0 However if similarity and knowledge con ict people tend to go with knowledge Demo reexamining classical view What is a bachelor 5year old boy Pope No Concept of eligibility we have a theory of what bachelors should be Summary of Conceptual Representations Classical Probabilistic Exemplar But similaritybased may not be enough need to take into account knowledge Betweencategory structure Hierarchy 0 Levels with setinclusion o 1 Tree maple tree sugar maple tree 0 2 Vehicle Car Sedan Property inheritance properties of higher levels inherited at lower levels 0 Properties of tree true of maple tree Taxonomic organization draw diagram example 9 Basic level Most people name objects in same way Basic level of representation common to people 0 Superordinate above Basic larger category furniture 0 Subordinate below Basic rocking chair Rosch s Explanations two opposing principles Informativeness 0 Want as much useful information as possible better for prediction explanation about that object 0 Get more information the further down you go 0 SO 9 lower is better Laziness 0 Want to use as little effort as possible 0 SO 9 higher is better Basic level is the best compromise Evidence for the basic level Superordinate levels are pretty much useless where as basic and subordinate levels are more useful subordinate is most useful Overlap of attributes feature listing Motor movements Shape similarity silhouettes Identi ed rst as this level Acquired rst Basic level can change with Context 0 If you re in a chair shop you have to ask for a type of chair not just chair 0 Expertise 0 Dog trainers are as fast on subordinates as on basics What makes a basic level Rosch compromise between maximizing information and minimizing effort information and laziness Highest level at which entities share parts Differentiation Differentiation Hypothesis two principles 1 Specificity maximize information want to get as speci c as possible 2 Distinctiveness do not want it confused with likely contrast categories categories at SAME level Superordinate not informative but distinctive Basic informative and distinctive Subordinate informative but not distinctive Basic level usually has the most distinctiveness but if items are more distinctive at the subordinate level expect that level to be faster than basic Superordinate example fruit vs vegetable most distinct Basic example apple vs orange vs very distinctive Subordinate example Macintosh apple vs Jonathan apple vs hard to differentiate at this low level Atypical subordinate example informative and distinctive have features that make them distinct from other subordinate categories Murphy and Brownwell 1985 Exp 3 Implications Basicness is not a property of entire level 0 Though there may be a level used for most typical items Why might distinctiveness be important 0 More accessible useful for prediction 0 Communication Complications Knowledge Basicness depends on expertise Basicness depends on measure 0 Can identify at one level e g bird 0 Reasoning implies basic at lower level robin 0 Why Types of conceptual organizations Taxonomic Thematic script items share a common association 0 Script based function of human activity Ad hoc category members that serve some goal or ideal Summary Conceptual representations classical probabilistic exemplar Theory view goes beyond similarity Between category structure 0 Hierarchies and basic levels 0 Explanation differentiation Type of conceptual representation 41 Problem Solving What is a problem 0 Givens initial conditions objects information 0 Goal desired outcome 0 Operators means of transforming conditions 0 Obstacles no simple direct known way from givens to goal The Tower of Hanoi Problem move stack to far right peg one disk at a time a disc can t be placed on a smaller disk Newell amp Simon approach Information processing system IPS Memory constraints 0 STM limits 0 Time to read out of LTM retrieve 0 Time to write into LTM encode Serial Processor basic thinking operations of problem solving is done on a serial basis comparing 2 items happens in a very serial fashion problem solving is an explicit deliberate conscious act Global program organization function of goals and knowledge are broken into parts Information Processing System IPS 9 Problem Space 6 Task Environment 0 Task Environment objective problem in the real world 0 Problem Space interaction of IPS and task environment Problem Space internal mental representation of the problem States description of problem elements 0 Initial state where we begin what we re given 0 Goal state where we want to get to 0 Intermediate states between initial and goal states Operators legal moves in space Apply operators to current state to try to reach the goal Search in problem solving Algorithm systematic procedure guaranteed to lead to correct solution 0 Examples addition maze 0 Problem computational complexity Heuristic strategy to guide search but not guaranteed to lead to correct solution 0 Examples hillclimbing working backwards 0 Problem no guaranteed solution Heuristics simple ones too simple always moving toward goal but sometimes need to plan Need to break up problem into meaningful parts subgoals and accomplish those Common heuristic Meansends analysis Basic idea look for difference between where you are in problem space initial state and where you want to be goal state try to eliminate the difference getting rid of big differences f1rst select operator to reduce difference Apply operator 9 new current state Problem Representation Main idea representation is crucial Representation can 0 Guide search for relevant knowledge 0 Determine which properties to focus on Maj or in uence on whether a problem can be solved How can a representation matter Must be through problem space States 0 Exclude o irrelevant parts OPERATORS Affect dif culty of applying operators Affect obviousness of some strategies What leads to representations Conventions Wording Expertise Schema Representation is a function of the input and activated knowledge understanding How to examine representation and search during problem solving Problem solving collect verbal protocols during solution talking aloud Computer program mimic the steps 0 Constrained by principles Analogy May underlie much of our learning Important part of problem solving transfer Correspondences Military Problem Tumor Problem Goal use army to capture fortress Goal use rays to destroy tumor Resources large army Resources powerful rays Operators divide army move army Operators adjust ray intensity move ray attack source administer rays warmy Constraints unable to safely use intense Constraints unable to safely send rays from one direction entire army along one road Solution Solution Outcome tumor destroyed by rays Outcome fortress captured by army Why is performance so low Not paying attention Do not understand the principle Did not think to use the General story So failure to notice Hypothesis contentdependence of knowledge Many different domains in LTM examples Why so hard People often seem driven by surface similarity 0 not so good transfer between tumor problem and military problem structural similarity o Other research shows that much better transfer between tumor problem and problem in which there is both structural and surface similarity What is the role of surface and structural similarity in accessing and using earlier examples How to test Vary surface and structural similarity of study item lightbulb to tumor problem Vary surface similarity vary instrument 0 laser vs ultrasound Vary structural similarity vary constraint 0 fragile glass vs insuf cient intensity Surface Cue Structural cues Laser Ultrasound F ragile glass High surface High structural Low surface High structural Insu icient intensity High surface Low Structural Low surface Low Structural PROCEDURE Read about ONE of 4 lightbulb story comp Write summary Given tumor problem to solve Testing No Hint Surface Cue Structural Laser Ultrasound Mean F ragileglass 54 Insu icient intensity 23 Mean 51 26 Testing Hint Surface Cue Structural Laser Ultrasound Mean Fragileglass 75 81 Insu icient intensity 60 47 Mean Implications Both similarity aid retrieval of appropriate source similarity helps people think of using analogous problem BUT similarity more important for drawing appropriate analogical inferences