The Role of Social Appraisal in Emotion

1
The Role of Social Appraisal in Emotion

• Tony Manstead
• Cardiff University

2
Overview

• What is social appraisal?
• Are our interpretations of emotional stimuli are
  influenced by the perceived reactions of others?
• Study 1 Videoconferencing paradigm
• Study 2 Co-viewing paradigm
• Are our expressions of emotion influenced by the
  anticipated reactions of others?
• Study 3 Gender differences in anger expression
  (vignette study)
• Study 4 Gender differences in anger expression
  (lab study)
• Conclusions

3
Emotion as a process Appraisal

• Appraisal is a term used to refer to the
  meaning an individual attaches to an event. As
  this meaning changes, so does the emotion (and
  if it changes, it must be a process)
• Appraisals are seen as reflecting the meaning of
  the event for the individual and its implications
  for his or her personal well-being
• But we often also appraise the emotional
  responses of others to the same event, and/or the
  anticipated consequences of ones own emotional
  response
• This is what we (Manstead Fischer, 2001) call
  social appraisal -- and it has deep roots

4
Social referencing
Gibson and Walk (1960)
Sorce et al. (1981)
5
Social appraisal

• the object of appraisal is frequently not
  just an event in which other persons are or are
  not directly involved behaviors, thoughts or
  feelings of one or more other persons in the
  emotional situation are appraised in addition to
  the appraisal of the event per se (Manstead
  Fischer, 2001, p. 222)
• So the core of social appraisals is the fact that
  people are sensitive to, and motivated to know,
  the emotional reactions of others

6
Two ways in which social appraisal influences
emotion

• Impact on the appraisal of an emotional event (as
  in the visual cliff situation)
• Others apparent or imagined interpretations of
  emotional events have an impact on how we
  interpret the same events
• Impact on the expression of emotion
• We anticipate how others are likely to react to
  our expressions of emotion and regulate those
  expressions accordingly

7
Impact on appraisalRelevant prior research

• Schachter (1959)
• Anxiety and affiliation The role of social
  comparison
• Schachter and Singer (1962)
• Two-factor theory of emotion The role of
  cognitions implied by others behavior
• Latané and Darley (1968)
• Where there is smoke Pluralistic ignorance
• Anderson, Keltner and John (2003)
• Emotional convergence hypothesis The role of
  contagion, empathy, and similarity in affective
  traits
• All of this hints at a process akin to social
  referencing in adults, but there is no conclusive
  evidence

8
Social transmission of face preferences among
humans
Jones, B. C. et al. (2007). Proc Roy Soc B
9
Present researchBruder Manstead (in prep.)

• Analogue of the visual cliff situation
• Confront adult participants with an emotional
  stimulus (film) in a setting where someone else
  is also exposed to that stimulus
• Vary how this other person reacts nonverbally to
  the film
• Assess participants own responses

10
Film stimuli
11
Study 1Overview

• 98 female students at University of Freiburg
  viewed film excerpts under different conditions
• confederate looked amused, disgusted or (blank
  screen) control
• Confederates facial reactions to films were
  pre-recorded and presented via (silent)
  video-conference system as if they were live
• Dependent variables
• Participants rated their subjective emotional
  responses and appraisals
• They were videotaped while watching the films,
  which allowed us to score their looking
  behaviour, and to code their facial behaviour
  (using a stripped down version of FACS)

12
Study 1Experimental set-up
13
Manipulation check for Study 1
14
Appraisal ratings for Stand by Me (film 1) and
Pink Flamingos (film 2)
15
Emotion ratings
16
Emotion ratings

• For both films, all the differences were in the
  expected direction, and in 3 of the 4 cases these
  differences were significant
• Ratings of amusement were higher in amusement
  condition than in disgust condition
• Ratings of disgust were higher in the disgust
  condition than in the amusement condition

17
Mediation analysis 1
Own appraisal
Film 1 .493 Film 2 .411
Film 1 .322 Film 2 .336
Experimental condition
Own disgust
Film 1 .043 (-.116) Film 2 .266 (.128)
18
Mediation analysis 2
Own appraisal
Film 1 .390 Film 2 .389
Film 1 .322 Film 2 .336
Experimental condition
Own amusement
Film 1 .244 (-.123) Film 2 .170 (.044)
19
Effects on facial behaviour
Only frequency of AU12 was significantly affected
by experimental condition
More smiles in social conditions than control
condition More smiles in amusement condition
than disgust condition
20
Study 1 summary

• Confederates nonverbal behaviour influenced
• participants appraisals of the films
• participants self-reported emotions
• The influence of confederates NV behaviour seems
  to be mediated by appraisals
• Some evidence of mimicking for AU12
• AU12 more frequent when confederate was amused
  than when she was disgusted

21
Study 2 Overview

• 56 same-sex dyads (friends or strangers) sat in
  front of a single monitor and viewed the same 12
  stimulus films together
• Although they were asked not to speak, they could
  otherwise interact freely in the partner
  visible condition
• In the partner not visible condition, a
  cardboard screen was placed between the dyad
  members

22
Study 2Experimental set-up
23
Assessing within-dyad similarity in response

• Our main hypothesis was that members of dyads
  would exhibit similarity in emotional reactions,
  and that this would be moderated by others
  identity and access to others NVB
• But how do you test this prediction?
• We used techniques described by Griffin and
  Gonzalez (1995)
• Key statistic is the pairwise intraclass
  correlation (rxx)
• It indexes the absolute similarity between two
  partners in a dyad on one outcome variable

24
Pairwise intraclass correlations for amusement
WHMS F F-, S S- SBM F S, S- S
25
Pairwise intraclass correlations for fear
TSh F F-, F S TSS S- S
26
Pairwise intraclass correlations for facial
amusement (AU12 ? AU6)
WHMS F S
27
Pairwise intraclass correlations for facial fear
(AU2 AU5 ? AU1 AU20 ? AU26)
TSh S S-
28
Pairwise intraclass correlations for frequency of
looking at other
29
Moderating effect of looking at dyad partner

• Dyads who scored high in their mean levels of
  looking at partner consistently converged more in
  their amusement levels (When Harry Met Sally,
  Stand by Me)

30
Study 2 summary

• Although the effects varied as a function of the
  film being viewed, dyad members -- especially
  friends who could see each other --had similar
  reactions to film excerpts
• This applied to self-report, facial behavior, and
  looking behavior
• High-looking dyads showed greater within-dyad
  similarity in emotion

31
Influence of social appraisal on the expression
of emotion

• We know that women and men differ in their
  expression of emotion
• Take the case of anger
• Although women and men generally do not differ in
  how angry they feel, there is evidence that men
  are more inclined to express their anger
• But the evidence is inconsistent
• Such inconsistencies may be due to the variation
  in social contexts that have been used in
  research
• Our assumption is that different social contexts
  evoke different social appraisals

32
Gender, anger, and social appraisal

• We also assume that men and women differ in their
  social appraisals
• Women tend to be more anxious about the possible
  negative consequences for others or for their
  relations with others when expressing anger
  (Eagly Steffen, 1986)
• Men tend not to anticipate such negative
  reactions, and may even expect positive outcomes,
  such as admiration (Campbell Muncer, 1987)
• Women are more likely to empathise with the anger
  target (Frodi Macaulay, 1977)

33
Evidence from own researchTimmers, Fischer,
Manstead (1998)

• Participants answered questions about vignettes
  describing emotional situations
• In pre-testing these vignettes were shown to
  elicit anger, fear, sadness, or disappointment
• Their task was to imagine that they were in the
  situations described, and to report
• how they would feel
• whether they would express emotion
• how they would express emotion
• their motives for expressing emotion

34
Timmers et al. (1998) Key findings

• Anger
• women more likely to cry or not show anything
• Fear
• women more likely to cry, say they feel afraid,
  freeze
• men more likely to not show anything
• Sadness
• women more likely to cry, say they feel sad, want
  to walk away
• men more likely to not show anything
• Disappointment
• women more likely to cry, say they feel
  disappointed
• men more likely to not show anything

35
Object-Target Relationship

• Object person who is the object of the emotion
• Target person to whom emotion is expressed
• Object-target relationship is same if
• I am angry at you and express my anger to you
• Object-target relationship is different if
• I am angry at you but express my anger to a 3rd
  person

36
Yelling and calling names when angry
Object-target relationship
37
Evers, Fischer, Rodriguez Mosquera, Manstead
(2005)

• Assumptions
• When women are angry, they tend to focus on the
  possible negative consequences of showing their
  anger -- restrict expression
• When men are angry, they tend to focus on the
  possible positive consequences of showing their
  anger -- express anger
• Objectives
• To show that women and men, despite feeling
  equally angry, express their anger differently
• To show that such differences are related to
  differences in social appraisal

38
Evers et al.Design and procedure

• Participants wrote essays that were evaluated
  by a fellow participant
• This other person provided positive or negative
  feedback about the essay
• Participants either expected or did not expect to
  meet this other person later
• In the context of a second, unrelated study,
  participant could choose how much hot sauce
  this other would have to consume (Lieberman et
  al., 1999)
• Amount of hot sauce serves as an implicit
  measure of anger expression

39
Evers et al. Anger induction

• Essay feedback was effective in evoking different
  emotions
• Negative feedback elicited more anger and sadness
• Positive feedback elicited more happiness and
  relief
• No gender differences, no interactions

40
Evers et al. Hot sauce allocation

• Amount of hot sauce administered was affected by
  feedback, gender, and social context
• Predicted 3-way interaction was significant

41
Evers et al. Mediation

• Was the gender difference in hot sauce allocation
  mediated by social appraisal?
• Mediational analysis

Negative social appraisals
.48
-.10
Gender
Hot sauce allocation
-.18
( - .15)
42
Evers et al. Conclusions

• Men and women were equally angry, yet women
  allocated less hot sauce in the social condition
• where participants believed that they were going
  to meet the object of their anger
• Negative social appraisals partly accounted for
  this difference between womens and mens anger
  expressions
• Women expressed anger to a lesser degree than men
  did because of their stronger expectations that
  angry behavior would have negative social
  implications

43
General conclusions

• The ways in which people experience and express
  emotion are shaped by social appraisals
• In Studies 1 and 2 we saw evidence that access to
  anothers nonverbal reactions to film excerpts
  influenced appraisals, self-reported emotions and
  facial behaviour
• In Studies 3 and 4 we saw evidence that gender
  differences in the expression of anger were
  apparent when the consequences of expressing that
  anger might damage social relations with others
• We pay attention to others emotions and take
  their reactions to our emotions into account
  because emotions play a central role in making
  social life possible
• Emotions provide us with the means to establish
  cooperative social relationships

44
The end

• Thank you