Regret, Norms, and Decision Process Quality

1
Regret, Norms, and Decision Process Quality

• Jochen Reb
• Terry Connolly
• University of Arizona
• December, 2004

2
Determinants of Regret Intensity

• Outcome severity
• (Bell, 1982 Loomes Sugden, 1982)
• 2) Outcome mutability
• (Kahneman Miller, 1986)
• 3) Decision justifiability
• (Connolly Zeelenberg, 2002)
• Many past results fit either OM or DJ account
  (e.g. action effect, abnormal behavior)
• ? This study compares the two explanations

3
Rival hypotheses

• 1. Mutability
• Regret more intense after abnormal action because
  its easier to imagine an upward counterfactual
  (If only Id done the normal thing)
• 2. Justifiability
• Regret more intense after unjustified action (If
  only Id done the proper thing, or done it the
  proper way)

4
Experimental Design
4 scenario-based experiments with student
participants
All factors manipulated between-subjects DPQ
decision process quality (careful or careless)
5
The Decision To V or not to V?

• Mother has to decide whether to vaccinate her
  small child against flu
• Bad outcome possible either way
• Vaccinate Maybe get bad side-effects of vaccine
  (kid sick for weeks)
• Dont vaccinate Maybe get flu (kid sick for
  weeks)
• Story says bad outcomes equally bad, and equally
  likely
• Independent variables
• Action vaccinate / dont vaccinate
• Personal normalcy make decision in usual /
  unusual way
• Social normalcy do / dont do what most people
  do
• Decision Process Quality make decision carefully
  / carelessly
• Dependent variable
• Regret when decision turns out badly

6
Results Personal Normalcy

• Main effect of Decision Process Quality
• -- less regret for careful dp
• Main effect of Action
• -- less regret for action

• Main effect of Action
• --- less regret for action

7
Results Social Normalcy

• Main effect of Decision Process Quality
• --- less regret for careful decision process

• Main effect of Action
• --- less regret for action

8
Discussion

• Results inconsistent with mutability
• No effects for normalcy in between-subjects
  analyses
• Action effects in opposite direction (more regret
  for not vaccinating)
• Results consistent with justification
• Less regret after good (justified) decision
  process
• Perceived justifiability mediates both
• process quality -gt regret link
• action -gt regret link
• Revised model

9
Future Research

• Do we need different justifications for different
  decision contexts? E.g.
• Careful, rational, cost/benefit type decision
  apparently justifies health-care decisions
• But would that process justify
• Partner decisions? (How much do I love thee, let
  me count)
• Ethical decisions? (Bribery showed the best
  cost/benefit ratio, so we did it)