Lecture Outline

1. Lecture Outline • Classical View of Schemas • Probabilistic View of Schemas • Types of Schemas • Models of Schemas • Hierarchical Levels of Abstraction • Associative Network Models • Parallel Constraint Satisfaction Model • Continuum Model of Impression Formation • Moderators of Schema Usage

2. Classical View There is a set of necessary and sufficient features needed for an instance to be categorized as a schema member • Schema members must have the necessary features AND • Having the necessary features is sufficient to be a schema member

3. Classical View Assumption 1: • If there is a necessary set of features to be a member, then schemas should have clear-cut boundaries Limitation 1: • Difficulty specifying defining features of instances (Wittgenstein) e.g., Games do not share a set of necessary features -- thus, boundaries are not clear

4. Classical View Assumption 2: • If there is a necessary set of features to be a member, then all members should be equally typical Limitation 2: • All members are not equally typical (Rosch) e.g., robins are more typical of birds than are penguins

5. Classical View Assumption 3: • If there is a necessary set of features to be a member, then categorization of new instances should be simple Limitation 3: • Some instances categorized more easily than others (Rosch) e.g., robins are categorized as birds more quickly than penguins

6. Probabilistic View No set of features is necessary or sufficient for an instance to be categorized as a schema member • Schema members share a family resemblance • Features that define schema membership form a fuzzy set -- none are necessary or sufficient

7. Probabilistic View of Schemas Addresses Limitations of Classical View Schemas do not have clear-cut boundaries • Without necessary or sufficient features to be a schema member: Boundaries of a schema are fuzzy--i.e., not clear-cut

8. Probabilistic View of Schemas Addresses Limitations of Classical View Schema members not all equally typical • Without necessary or sufficient features to be a schema member: Some members may be more typical than others

9. Probabilistic View of Schemas Addresses Limitations of Classical View Categorization of new instances not simple • Without necessary or sufficient features to be a schema member: Typical instances should be categorized more easily than atypical instances

10. Person Schemas Who are you? • Traits that co-occur in others • Extroverts are outgoing and friendly • Introverts are quite and shy • Behaviors that characterize person-types • Extroverts like big parties • Introverts like small gatherings Function: Help people draw inferences about others

11. Self Schemas Who am I? • Everything one knows and can imagine about oneself • Traits characteristic of oneself • Memory of one’s past • Expectations for one’s future self Function: Help organize, guide, and interpret incoming information

12. Self-Schemas Schematic: clear self-view on a dimension • important and central to one’s self-concept Aschematic: unclear self-view on a dimension • not important and not central to one’s self-concept

13. Self-Schemas: Information Processing Markus (1977) Purpose: Test whether self-schemas help people process information Prediction: People will process information more quickly when they are schematic than aschematic on a dimension

14. Self-Schemas: Information Processing Markus (1977) Procedure: • Assessed whether participants were schematic or aschematic on independence/dependence • Schematics: extreme (in)dependence • Aschematics: moderate (in)dependence • Participants indicated as quickly as possible whether a series of traits described them • Expectations about typical behavior • Extroverts talk a lot at parties

15. Markus (1977) Results: • Schematic-Independents responded faster to independent than dependent traits • Schematic-dependents responded faster to dependent than independent traits • Aschematics responded similarly to independent and independent traits • Conclusion: self-schemas enable one to process self-relevant information more quickly

16. Role Schemas What are those people like? • Norms and expectations for particular roles in society • Waitresses take food orders • Doctors cure the ill Function: • Help people draw inferences about others • Simplify social information

17. Role Schemas Achieved roles: • typically acquired through effort and training • pro-basketball player • college student Ascribed roles: • typically acquired through birth • gender • ethnicity

18. Event Schemas What happens here? • Expected sequence of events • going to class • going to the gym Function: • Help people anticipate what happens next • Help people achieve next step in sequence via planning and goal setting

19. Models of Schemas Hierarchical: Levels of Abstraction • Schemas organized in hierarchies Example • Human Being • Female • Individual

20. Cantor & Mischel (1979) Prediction: Schemas have 3 levels of abstraction • superordinate (human being) • middle (female) • subordinate (individual)

21. Cantor & Mischel (1979) Emotionally Unstable person Superordinate Phobic Criminal Madman Middle Claustro- phobic Hydro- phobic Acro- phobic Strangler Torturer Rapist Subordinate

22. Cantor & Mischel (1979) 4 Hierarchies • Emotionally unstable person • Cultured person • Extroverted person • Committed (to belief or cause) person

23. Cantor & Mischel (1979) Procedures: 1.Gave people 32 cards with middle and subordinate labels 2.Instructed participants to sort cards into the 4 superordinate categories 3. Participants subdivided piles into less inclusive piles

24. Cantor & Mischel (1979) Result: Participants sorted the cards into three levels of abstraction Participants’ piles looked like one shown before

25. Cantor & Mischel (1979) Richness: Number of features Differentiation: Overlap between categories Vividness: Detail of features

26. Cantor & Mischel (1979) Richness • Participants given one level of abstraction Either: 4 superordinate levels 2 middle levels w/i taxonomy 6 subordinate levels w/i taxonomy

27. Cantor & Mischel (1979) Richness • Participants given 2.5 minutes to list typical features of the level given • Idiosyncratic features removed

28. Richness Number of Features Listed Hierarchy: Superordinate Middle Subordinate 12 20 18 Cultured 12 25 32 Extroverted 8 17 26 Committed Emotionally Unstable 10 22 22 Result: Hierarchies became richer as they became more specific

29. Cantor & Mischel (1979) Differentiation • Examined overlap in listed features across levels of abstraction Example: Overlap in features for • phobic and criminal madman • emotionally unstable person, cultured person, extraverted person, committed person

30. Differentiation Amount of Overlap Hierarchy: Superordinate Middle Subordinate 0 9 11 Cultured 0 8 23 Extroverted 0 4 16 Committed Emotionally Unstable 0 2 20 Result: Hierarchies became less differ-entiated as they became more specific

31. Cantor & Mischel (1979) Vividness • Examined how detailed features were at each level. Focused on: • physical features or possessions • socioeconomic status (SES) • traits • behaviors

32. Vividness Number of Features Listed Middle Hierarchy: Superordinate Subordinate .50 1.50 2.33 Physical/Possession .25 .38 .54 SES 8.75 15.75 17.21 Traits Behaviors 1.50 3.88 4.75 Result: For each kind of feature, number increased as level of abstraction became more specific

33. Cantor & Mischel (1979) Summary • Richness and vividness increased as levels became more specific • Differentiation decreased as levels became more specific

34. Models of Schemas Associative Network Models • Schemas organized as web of features • Nodes = features • Links = association between features

35. Associative Network Models Protests unfair treatment Wants nice house Won’t pay rent until house painted Aggressive Curses Punches Hits Asian American Lawyer Well dressed Competitive Intelligent Link Node

36. Activation of Nodes Context affects a node’s level of activation e.g., surfing v.s. when will class end

37. Activation of Nodes Adjacent nodes activate each other (Called Spreading Activation)

38. Activation of Nodes Nodes can be simultaneously activated by multiple other nodes

39. Activation of Nodes More activation = node has more affect on processing (memory, inferences)

40. Activation of Nodes Activation decays gradually

41. Associative Network Models Limitation: Activation continues indefinitely Here’s how……….

42. Associative Network Models Protests unfair treatment Wants nice house Won’t pay rent until house painted Aggressive Curses Punches Hits Asian American Lawyer Well dressed Competitive Intelligent

43. Associative Network Models BUT…… Assertive is not associated with Asians Thus, model breaks down

44. Parallel Constraint Satisfaction Models Same as Associative Network Models except: • Excitatory AND inhibitory links • Positive AND negative links

45. Parallel Constraint Satisfaction Models Excitatory Links: • Nodes activate each other • Aggressive activates Lawyer Inhibitory Links: • Nodes deactivate each other • Asian American deactivates aggressive

46. Parallel Constraint Satisfaction Models Positive Links: • Nodes both activated or deactivated • When Aggressive activated, Lawyer activated • When Aggressive deactivated, Lawyer deactivated Negative Links: • One node activated, one deactivated • When Asian American activated, Aggressive deactivated • When Asian American deactivated, Aggressive activated

47. Models of Schemas Continuum Model of Impression Formation (Fiske & Neuberg, 1991) • Dominant model in field right now • Explains how people use role schemas to form impressions (See page 136-139 in text)

48. Continuum Model: Main Ideas • People use schemas to conserve mental effort (attention) • Impression formation is a continuum of processes • Each process requires more mental effort (attention) than the one before it • Each process reflects less influence of schema than one before it

49. Points On the Continuum Point 1: Initial Categorization • Categorize target • Warrant further processing? • Stop processing and base impression on schema OR move to next point

50. Points On the Continuum Point 2: Confirmatory Categorization • Match target to category • If match good: • stop processing • Use schema to form impression • If match poor: • allocate more attention to person • move to next point