Methods for Assessing Social Referencing of Risk Perception

1
Methods for Assessing Social Referencing of Risk
Perception

• Gwenda Simons Brian Parkinson
• Department of Experimental Psychology
• University of Oxford

2
Background

• Understanding how people arrive at choices where
  risk is involved often focuses on the individual
  in isolation.
• Previous research shows that own affect
  influences perceptions of decision options
• Affect priming (automatic activation of
  associated semantic content).
• Affect as information (how do I feel about it?
  Schwarz Clore, 1983).

3

• In our research project we consider if and how
  people may take into account the (emotional)
  responses of another person in reacting to- and
  dealing with worrying issues or decisions
  involving some level of risk or uncertainty.
• Influence of another persons affect may occur by
  modulating our appraisal of the (decision)
  situation.
• The reactions of another person can help us to
  make sense of a situation or give us an idea of
  how the other person appraises the situation and
  the risk involved and thus influences our
  behaviour.
• In turn our affective reactions will influence
  their appraisals and emotional reactions.

4

• We see emotions not as individual responses to
  private appraisals, but as interpersonal
  processes of relational alignment.
• Emotional communications are oriented to actual,
  perceived, and anticipated (emotional) responses
  of others.
• Affect is not simply an informational input to a
  decision process, but part of the process of
  co-ordinating perspectives on an issue (How do
  you feel about it? How do you feel about how I
  feel about it? etc).

5

• The research project uses different and
  innovative methods
• Diary studies
• Interviews
• Laboratory experiments (quiz-game)
• Online survey (Risk Inventory)
• Diary and observation of real-time interactions
  ( video-cued recall)

6

• This interaction research is collaborative
• Claudia Marinetti (DPhil candidate researching
• the Illusion of Transparency)
• We have further received assistance from the
  following internship students
• Michaela Rohr, Kornelia Gentsch,
• Robyn Migliorini Mina Yadegar

7
Interaction study Research questions

• How do people regulate their own and each others
  emotions when interacting with their partner in
  relation to issues which worry them or pose a
  certain amount of risk?
• Does such regulation always have its intended
  effects?
• Can we detect specific patterns of interpersonal
  emotion regulation in people sharing a close
  relation? (e.g. one partner always expressing his
  or her worry or anxiety or indeed enhancing it
  when the other partner does not).

8
Method

• 16 couples with at least one child under the age
  of ten (14 mixed gender, 2 same gender).
• Aged between 25 and 57 (M 36.03, SD 7.36).
• Mixed background and occupation

9
Diary keeping

• Each partner provides diary data over the course
  of 1-2 weeks for every discussion the couple has
  about concerns/decisions relating to their young
  child(ren)
• Description of the concern/decision
• Appraisals of these concerns
• Emotional reactions to these concerns
• Attempts to regulate their own emotion
• presentation and that of their partner. 

10
(No Transcript)
11
Lab-based session

• Partners are video-recorded whilst talking about
  some of the issues recorded in their diaries.
• We select 2 topics
• 1 where the partners agree on their level of
  worry experienced during the original discussion
• 1 where the partners disagree on their level of
  worry
• Several questionnaires are completed (Interaction
  Rating Scale, BEQ, ERQ, TAS).
• Computer-based emotion recognition task.
• Video-cued recall procedures (e.g., Levenson
  Gottman, 1985).

12

• Video-cued recall procedure in order to solicit
  reports of
• Experienced level of worry.
• Emotion regulation (own and others emotions).
• Perceived emotion and emotion regulation in the
  other partner.
• Facial expression.

13
Some preliminary findings

• We split the diary topics into those where the
  partners agreed on the level of worry they
  experienced in relation to the discussion and
  those where they disagreed.
• Worry agreement diary topics
• Minimal or no difference in level
• experienced worry between partners
• Overall worry level (M 3.81, SD 1.59)
• Worry disagreement diary topics
• One partner more worried than the other
• (M difference 2.75, SD 1.12)
• Overall worry level (M 4.00, SD 1.85)

14
Examples of issues discussed

• Whether or not to take son to see doctor.
• Whether or not to follow health visitors
  suggestion to wake baby daughter up during the
  night for a feed because she is not putting on
  enough weight.
• What to do about childs misbehaviour in
  nursery.
• Son is being bullied at school.
• Daughter is grinding her teeth at night.

15

• Participants reported that the lab-based
  conversations felt fairly natural (M 4.33,
• SD 1.54 on 7-point scale).
• Levels of worry recorded in the diary and the
  mean levels of worry experienced in the lab-based
  conversation (as measured in the video-cued
  recall) correlated positively and significantly
  (r (64) .346, p .005).

16
Video-cued recall data

• Initial analysis of the Video-cued recall data
  revealed some interesting findings
• Around half of the couples showed clear
  disagreement in the level of worry that one
  partner experienced and the other partner
  perceived.
• The greater the difference in experienced worry
  (i.e. the worry reported by both partners), the
  less successful one partner was in correctly
  identifying the level of worry in the other
  (correlation absolute difference scores r(64)
  .819, p lt.001).

17

• Video-cued recall data continued
• Difficulties in correctly identifying their
  partner's level of worry may be due to
  difficulties in detecting the level to which
  their partner is actually trying to hide or
  enhance his or her expression of worry
  (correlation absolute difference scores r(64)
  .387, p .002).
• Overall partner B (in all but 2 cases the male
  partner) was less worried by the issues discussed
  in the lab compared to partner A (always female),
  t(62) 2.12, p .038.
• Partner A perceived her partner to be somewhat
  more worried than he (or she) actually felt (M
  difference -0.95, SD 1.42), whereas Partner B
  perceived his (or her) partner to be a little
  less worried than she actually felt (M difference
  0.42, SD 1.60).

18
Some early conclusions

• We were able to recreate real-life interactions
  in the lab.
• Video-cued recall was a successful method for
  discovering the patterns of emotional reactions
  to the worrying (and risky) issues reported in
  the diary and discussed in the lab.
• There are indications that those partners who are
  aware of their partners attempts at emotion
  regulation are better at detecting their
  partners true levels of experienced emotion
  (and act accordingly?!).

19

• More analyses to follow
• More elaborate analysis of the reciprocal
  patterns of emotional influence.
• Specific factors which influence the patterns
  (e.g. ability to recognise emotional expressions,
  facial expressiveness, tendency to inhibit
  expressions).
• Speech content.
• Analyses of nonverbal behaviour (FACS and
  untrained judges).

20
Thank you!
Contact gwenda.simons_at_psy.ox.ac.uk
21
ltback