Presentation is loading. Please wait.

Presentation is loading. Please wait.

Homework Review Participants given either an injection of epinephrine (which increases heart rate) or a placebo (salt water) Participants were next assigned.

Similar presentations


Presentation on theme: "Homework Review Participants given either an injection of epinephrine (which increases heart rate) or a placebo (salt water) Participants were next assigned."— Presentation transcript:

1 Homework Review Participants given either an injection of epinephrine (which increases heart rate) or a placebo (salt water) Participants were next assigned to ‘happy condition’ (silly tasks) or ‘angry condition’ (annoying tasks) Result: Participants receiving epinephrine reported – more happiness, happy behaviors in ‘happy condition’ compared to those receiving the placebo. – more anger, angry behaviors in ‘angry condition’ compared with those receiving the placebo.

2 “Additive” Interpretation AmuseAnnoy Annoyed Negative Positive High Arousal Low Arousal Amused

3 “Additive” Interpretation AmuseAnnoyDrug Angry Negative Positive High Arousal Low Arousal Happy

4 Schacter & Singer Claim we don't automatically know when we are happy, angry. Instead, we label our emotions by considering situational cues. This labeling process depends on two factors: – something in world triggers general, nonspecific arousal marked by increased heart rate, tightening of the stomach, and rapid breathing. – people examine perceptual cues and their thoughts about the situation for clues for what has caused the emotion. Essentially 2 factor model – Appraisal a purely cognitive process that provides context for interpreting body state

5 “Emotion” Elicitor Model BIS/BAS Mental State Elicitor Emotion Model Mental State Elicitor Emotion Model Discrete DimensionalAppraisal

6 Another example Dutton & P. Aron had male participants walk across either a scary (arousing) or safe suspension bridge. At the end of each bridge an attractive female experimenter met the participants, gave them a survey and a number to call if they had any other further questions. Men who walked across the scary bridge were most likely to call the woman for a date CRITICISMS: Many failures to replicate original Schacter & Singer experiment Vague with regard to role of cognitive mechanism Ecological validity: Most situations the arousal comes from the situation. So may be true but only part of the story

7 Social Emotions New experiment

8 Emotion influences your demains Celso M. de Melo, Peter Carnevale & Jonathan Gratch No main effect of Modality (p=.383>.05) – Participants don’t concede differently, when collapsing across modalities Main effect of Emotion (p<.01) – People concede more to the angry agent than the control agent – People concede less to the happy agent than the control agent No interaction between Emotion and Modality (p=.550>.05) – Suggests emotion is having same impact independently of modality

9 Why?

10 Social Emotions Discuss view that emotions arise from and influence social interactions and relationships –Framework for analyzing emotional “transmissions” –Controversies about origins of emotional behavior –Function of emotional signals –Social appraisal theory

11 Emotional communication Commonsense view: –Something leads a person to experience an emotion –Emotion alters the senders behavior –Observers notice and interpret this behavior –This interpretation changes the observer’s behavior Anger Fear

12 Transmission model Shannon-Weaver (1947) Transmission Model of Communication SAD MAD

13 Transmission model Basic transmission questions –What “message” is being encoded/decoded Basic emotion (Fear, Hope, Joy) Appraisal (Uncertainty, Goal Congruence) Broad dimensions of affect (valence, arousal) –How is the message encoded into behavior what features are diagnostic of true emotional state? –How is message decoded What features to people use to infer emotion? Are they any good at it? Are they the correct features Secondary social questions –Why was message generated True emotion? Strategic concerns? –How does it impact observer behavior?

14 SenderReceiver Brunswik’s Lens Model Affective State Inferred Affective State Behavior Agent

15 Attend to User Smile Nod Recline Extend Leg Signal Challenge for Detection: Encoding/Decoding Mismatch Brunswik’s Lens Model Sender Induce Affective State Receiver Decoding Check if state “Received”

16 Encoder SenderReceiver Speech Prosody Expression Body Lang. EEG Decoding Affective State Affective State Signal Brunswik’s Lens Model Decoder

17 Encoder SenderReceiver Speech Prosody Expression Body Lang. EEG Decoding Affective State Affective State Signal Brunswik’s Lens Model Decoder Encoding Model Decoding Model =

18 Naturalistic Example Can you decode this

19 Evidence for transmission model of Emotion ENCODING Is “true emotion” encoded into body movements? –Evidence mixed (Russell et al 03, Feldman-Barrett) –No consensus that “Emotion” is encoded –More consensus that “core affect” is encoded Facial & vocal changes coordinated w/ sender valence and arousal

20 Decoding: People think they can do it well Emotions are complex and dynamic This actually matters significantly for the observer

21 Attend to User Smile Nod Recline Extend Leg Signal Challenge for Detection: Encoding/Decoding Mismatch How would we test? Encoder Sender Warm – Agreeable Gifford (1994)

22 Encoder Sender Attend to User Smile Nod Recline Extend Leg Warm – Agreeable Signal Gifford (1994) Challenge for Detection: Encoding/Decoding Mismatch Some affective messages reliably encoded

23 Receiver Attend to User Smile Nod Recline Extend Leg Decoding Decoder Inferred Warmth Signal Challenge for Detection: Encoding/Decoding Mismatch But not always reliably decoded

24 ReceiverSignal Encoder Sender Attend to User Smile Nod Recline Extend Leg Decoding Decoder Warm – Agreeable Inferred Warmth Gifford (1994) Challenge for Detection: Encoding/Decoding Mismatch Encoding/Decoding Mismatch ≠

25 Evidence for transmission model of Emotion DECODING Is “true emotion” decoded –Evidence mixed –Most people can infer something of the sender’s psychological state Voice: arousal, not valence (high confusion between happy/angry) Face: valence and arousal Smiles universally recognized across cultures, not much else Decoding depends on context, gender of sender, receivers emotional state *But what do psychologists know about signal processing

26 ReceiverSignal Encoder Sender Attend to User Smile Nod Recline Extend Leg Decoding Decoder Warm – Agreeable Inferred Warmth Gifford (1994) Challenge for Detection: Encoding/Decoding Mismatch Question: If Encoding ≠ Decoding what are implications for affective computing? ≠

27 Signal Encoder Sender Attend to User Smile Nod Recline Extend Leg Warm – Agreeable Gifford (1994) Challenge for Detection: Encoding/Decoding Mismatch Question: If Encoding ≠ Decoding what are implications for affective computing? Suggests that people attend to wrong cues Maybe computers could do better at detection? Maybe computers could inform or teach people to do better

28 ReceiverSignal Attend to User Smile Nod Recline Extend Leg Decoding Decoder Inferred Warmth Challenge for Detection: Encoding/Decoding Mismatch Question: If Encoding ≠ Decoding what are implications for affective computing? Suggests that people attend to wrong cues Maybe computers should give people what they want Use displays that meet user expectations Some evidence that this is what actors do

29 Returning to people: Do we have to abandon lens model? One explanation: No but it’s complicated –People express true emotion but are motivated to mask their displays for social reasons –Ekman’s Display rules –If you look close enough, you’ll see evidence of the true emotion (Microexpressions)

30 “Leakage view” Is “true emotion” encoded into body movements?

31 Alternative explanations: Transmission model fine –BUT message isn’t about emotion People don’t reveal true emotion People attempt to influence others Emotional displays are strategic messages

32 Social Intention view Fridlund –It is evolutionarily stupid to reveal true feelings Gives others too much information We often need to cheat, lie, seal –Little evidence for direct mapping between emotion and display Many to 1: Laugher produced by humor, anger anxiety, self-deprecation, attention, appeasement, sexual interest 1 to Many: Ekman and Friesen78 identify 65 anger displays Disassociation: –Happiness neither necessary nor sufficient for smiling (Kraut&Johnson79) –Surprise not correlated with “surprise” display (Reisenzein) –Emotional displays vary depending on the social context Tend to smile more when someone watching –Emotional displays are strategically deployed Show distress when someone that could help us is near

33 Break?

34 Social Functional view of emotion Emotion plays crucial interpersonal function to facilitate coordinated activities: e.g. conversation –Hypothesizes people are social by nature and survive through relationships (Lutz&White, 1986). –Hypothesizes emotion is a feeling system: Provides intrinsic rewards for cooperation Provides intrinsic punishments for non-cooperation –Hypothesizes emotion is a display system Automatic and rapid means of communicating mental state Facilitates detection of cheaters –Hypothesizes tight coupling between feeling and display You display what you feel Point of contrast with social intentions views (Keltner & Haidt 99; Levenson; Isard; Frank)

35 Social function of felt emotion

36 Provides information about social encounters –Informs self about quality of the interaction (Keltner&Haidt99; Clore05) Anger: fairness of events Love: level of commitment to another Shame: lowered social status

37 Social function of felt emotion Provides information about social encounters Prepares the body for social responses –Prepares response to social contingencies –Eg: Anger (Keltner & Haidt 99) Shifts blood from internal organs towards the hands and arms Heightens sensitivity to injustices of others which presumably Thereby facilitates responses to threat or injustice.

38 Social function of felt emotion Provides information about social encounters Prepares the body for social responses Motivates formation of group bonds, loyalty, identity –Trusting others feels good Acts of trust correlated with elevated levels of oxytocin, same chemical released during breast-feeding (Zak, 2004) –Harming others feels bad Anticipatory guilt and shame and help enforce social rules (Barrett 1995, Izard et all 98) –Discrepancies between felt emotion and other’s responses stimulates development of perspective-taking (Dunn95), theory of mind (Harris89), self (Eder90)

39 Social function of displayed emotion

40 Elicits adaptive social responses from others –Anger Elicits fear-related responses (even subliminal presentation) (Dimberg&Ohman96) Serves as demand for someone to change course of interaction (Emde, Gaensbaur&Harmon76) Diverts blame or averts someone else’s anger or disapproval (cf. Biglan et al., 1985; Clark, Pataki and Carver, 1996). –Distress elicits sympathy (Eisenberg et al89) –Joy foster social interaction (Haviland&Lelwica87)

41 Social function of displayed emotion Elicits adaptive social responses from others Transmit coordinating information (Spoor&Kelly04, Parkinson01) e.g. social referencing –Beliefs, desires & intentions –Appraisals: desirability, control, expectedness –Orientations towards the relationship (dominant/submissive) –Not strictly a transmission model Can emerge through dyadic interaction. Eg. antagonism may develop as facial expressions and bodily postures are exchanged: One person’s leaning forward leads to withdrawal until ground is held. (Parkinson 2001)

42 Social function of displayed emotion

43 Social Emotional displays in Primates From de Waal, F. (2006). Primates and Philosophers: How Morality Evolved. Princeton, NJ, Princeton University Press

44 Emotions motivate dialogue toward shared meaning –Seek social support (Carver, Weintraub, & Scheier, 1989; Stober, 2004; Dukel-Schetter, Folkman, & Lazarus, 1987) –As part of a need to “search for meaning” (Luminet, et al., 2000; Rimé, Mesquita, Philippot, & Boca, 1992) –As part of a “reality negotiation” to mitigate the negative impact of a possible transgression, through the use of excuses (Snyder, 1989; Snyder & Higgins, 1997) Conversational function

45 (act on world)(act on self) Emotion Coping Strategy Action Tendencies “Affect” Physiological Response Problem-FocusedEmotion-Focused Environment Goals/Beliefs/ Intentions Theoretical Framework: Appraisal Theory (Arnold, Lazarus, Frijda, Scherer, Ortony et al.) Desirable Expected Controllable Blame Appraisal

46  Past Present Future  Goal: Establish trust Utility: 50 Action: Agent Defects Cause: Self Intend: yes Probability: 100% Appraisal Desirability: -50 Likelihood: 100% Causal Attribution: Self Emotion: Guilt(50) Inhibits How does an agent know what to feel? Threatens

47  Past Present Future  Goal: Establish trust Utility: 50 Action: Agent Defects Cause: Self Intend: yes Probability: 100% Appraisal Desirability: -50 Likelihood: 100% Causal Attribution: Self Emotion: Guilt(50) Inhibits And how do these “feelings” influence people? Threatens Hareli, S. and Hess, U. What emotional reactions can tell us about the nature of others: An appraisal perspective on person perception. Cognition & Emotion 24 (2010), 1, 128 — 140 “Reverse Appraisal”

48  Hypotheses: –Meaning of expression is context-dependent (requires reverse appraisal)  It doesn’t matter that they smile, it matters what smile communicates about person-environment relationships  Expression interpretation depends on action and (unobservable) goals  To test: –Use appraisal theory to generate expressions of a ‘confederate’ –See if people accurately “decode” this meaning (via reverse appraisal) –See if people predictably alter their behavior based on recovered meaning Testing reverse appraisal

49  Created game-playing agent: Iterated Prisoner’s Dilemma Multi-issue bargaining  From appraisal theory: Give different agents different goals –e.g., “cooperative” agent has goal to foster joint reward –Appraise state of game –Display emotions consistent with goals  Assess effect (if any) on human behavior Social Effects of Agent Emotions

50  Cooperative vs. Individualistic Individualistic agent Agents express similar emotions but elicited under different contexts Shows joy when agent defects on human (51 participants, within-subjects design)  Play 25 rounds IPD for chance at $100  Participants believe they are playing against computer  Also explore several agent appearances  Predict greater cooperation with cooperative agent Experiment 1: Iterated Prisoner’s Dilemma

51 Expression Policy “Cooperative” agent “Individualistic” Agent CooperateDefectCooperateDefect Coop JoyGuilt Coop NeutralJoy Def AngerSad Def Sad Human Same expressions, only difference is situation in which they are evoked (consistent with different appraisals)

52 Example interaction: Iterated prisoner’s dilemma

53 Results: 2 (expression-policy) x 3 (appearance) design: 51 participants  “Cooperative” emotions –Enhance cooperation –Enhance perceptions of trust  “Competitive” emotions –Increase perceptions of dominance –Lower perceptions of trust  Identical expression has different effects depending on situation in which it is evoked –i.e., contingent meaning

54 Real World Example (Golden Balls)

55 Step Back a second “Emotion” Elicitor Model BIS/BAS Mental State Elicitor Emotion Model Mental State Elicitor Emotion Model Discrete DimensionalAppraisal

56 Mental State World Emotion Model R WHAT IS SOCICAL FUNCTION OF EMOTION: Psychological Model General consensus: Models suggest some relationship (correlational or causal between Mental state (beliefs, desires, intentions) The world (actions, other actors) Emotion (including behavior) But what good is this Challenges Methodology

57 MSW E Model R Psychological Model MSW E Model F Folk Model Model R ≈ Model F ? Distinction that may be important: “real” vs. “naïve” model Accurately reflects how emotion really happens or is displayed (i.e., encoding) Accurately reflects how people THINK emotion really happens or is displayed (i.e., decoding)

58 MSWorld Emotion? Model R Infer emotional reaction: If know Goals and Action: guess response Facial expression Risky decision Run away MS?World Emotion Model R MSWorld? Emotion Model R Infer mental state If know Response, Action, guess goals cooperative vs. competitive Low reservation price Infer unobservable actions If know goals and reaction, guess action Lion in the grass INFERENCE : Enabling AGENT inferences about HUMAN (Can do this regardless of assumptions about causality: correlation sufficient)

59 MSWorld Emotion? Model F Infer emotional reaction Facial expression Risky decision Run away MS?World Emotion Model F MSWorld? Emotion Model F Infer mental state cooperative vs. competitive Low reservation price Infer unobservable actions Lion in the grass INFERENCE : Enabling HUMAN inferences about AGENT

60 Influence MS via emotion - sad induction leads to belief that hill is high SOCIAL PREDICTION : Form AGENT expectations over future HUMAN beh. Induce Sadness Mahler MS(t)W(t) E(t) Model R ∆MS(t+1)W(t+1) ∆E(t+1) Model R

61 SOCIAL PREDICTION : Form HUMAN expectations over future AGENT beh.

62 Influence MS via emotion - sad induction leads to belief that hill is high SOCIAL PREDICTION : Form AGENT expectations over future HUMAN beh. Induce Sadness Mahler MS(t)W(t) E(t) Model R ∆MS(t+1)W(t+1) ∆E(t+1) Model R

63 MSW Emotion Model R Influence human emotion via MS -tell human “I cheated” to provoke anger MSW Emotion Model R SOCIAL ACTION : Facilitate agent control over human emotion (here causality matters) Influence human emotion via W -hit human to provoke anger speech act or Expression Action in world

64 Influence MS via emotion - sad induction leads to belief that hill is high SOCIAL ACTION : Facilitate agent control over human MS via Emotion MSWorld Emotion Model R Induce Sadness MS(t+1)World Emotion Model R Influence their action (get them to take risky choice by anger induction Model Function: Facilitate agent control over human Action Mahler Induce Anger

65 SOCIAL ACTION : Construct emotional plans MSW E Model R MSW Sad Model R MSGive Up Sad Model R

66 SOCIAL ACTION : Facilitate human control over agent

67 Challenges  Need the model –which way do the arrows point –do we need more boxes –This determines reliability and scope of  explanation  prediction  control


Download ppt "Homework Review Participants given either an injection of epinephrine (which increases heart rate) or a placebo (salt water) Participants were next assigned."

Similar presentations


Ads by Google