If consistency is high, and distinctiveness / consensus are low, then a personal attribution is more likely:
“The baby is always smiling, never displays other emotions (like crying), and this is not typical of babies in general. Therefore, this baby must have a happy disposition.”
If consistency is high, and distinctiveness / consensus are also high, then a situational attribution is more likely.
“The baby is always smiling when tickled, but displays different emotions in other circumstances. Smiling when tickled is typical of all babies. Therefore, this baby is smiling because it was tickled”
A researcher assigned participants to read out loud either a pro-Castro essay or an anti-Castro essay. A group of listeners rated the extent to which the reader held pro-Castro or anti-Castro beliefs.
Even though the listeners knew that the readers had no choice in which essay to read, the raters judged the pro-Castro readers as being more pro-Castro than the anti-Castro readers.
The listeners failed to take into account the strong situational factor present (that the readers had no choice about which essay to read).
Choice leads to stronger attributions of liking.
Cognitive dissonance theory
People strive for consistency in their thoughts.
Seems simple, but this idea lead to very counter-intuitive findings.
Festinger and Carlsmith had participants complete a very boring task (turning screws ¼ turn at a time, for a long time). One group was paid $1 to do this, and a separate group was paid $20. Which group should like the task more?
Participants paid $1 enjoyed the task more than participants paid $20. The $20 group had consistent cognitions (“This is stupid, but I’m being compensated for doing it.”). The $1 group had inconsistent cognitions (“This is stupid, and I have no reason to do it.”).
Participants in the $1 group resolved the dissonance by changing their attitude about the task (“I’m not being adequately compensated for this, but that’s OK. This is actually fun!”).
Bem suggested that another way to think about this research is in terms of attribution.
All participants observed their behavior (doing the boring task) and made a causal attribution for their own behavior. Participants in the $20 group observed their own behavior and thought (“I’m doing this task because I’m getting paid.”). Participants in the $1 group observed their own behavior and thought (“I’m doing this task because I must enjoy it.”).
"By altering actors and observers perspectives through videotape replays, mirrors, or other methods, one can correspondingly alter the actors' and observers' causal assessments."