Dr. Grossi's Blog

Actor-Observer Bias

Dr. Philip Grossi
Monday, 02 May 2011

In 1971, E.E.Jones and R.E. Nisbett wrote a book about the different perceptions of the causes of behavior.  The popular view is that people tend to view their own actions as caused by the situation and the behavior of other to result from their personalities.  This view is a bit oversimple.  Individuals are highly aware of the of the emotions produced by situations, and it is these feelings that are the core of their attributions and not situations that automatically control their actions.  On the other hand, other's actions are viewed as constrained by what is viewed as stable personalities.  Therefore, one's own actions are viewed as more reflective of free will that that of others.

illustration to actor observer bias blog

What is free will?  It implies the ability to choose from alternatives and that those alternatives are therefore not determined.  This implies that a person can take other actions, and furthermore, that the course of action taken is a reflection of his/her intentions  or desires.

Now comes a report of four experiments which shed light on these issues and which are centered on predictability (Experiment 1),possibilities (Experiment 2), possibilities and self-enhancement (Experiment 3), and agency (Experiment 4). These experiments were carried out on Princeton students or workers in a pizza parlor in Princeton, New Jersey.


Experiment 1:Hypothesis: The participant's past and future decisions would be less predictable than those of their roommates.  There was no self-other differences between past and future.  The results supported the hypothesis that the participants' decisions regarding their life would be less predictable than those of their roommates about whom they had a reasonable amount of information.

Experiment 2: Hypothesis: If free will exists, then many possibilities must exist from which an individual can choose, and that the future of participants (residence, lifestyle, job) should contain more possibilities than the future of others.  Indeed participants circled more options that were considered genuine possibilities for themselves than for their co-workers.  An additional prediction was that participants (waiters in an Italian restaurant in Princeton, New Jersey)  would see their co-workers as in the same position as today 10 years from now.  Indeed that was endorsed more for the co-worker than for the participant.  The answers were also rated in terms of most desirable and least desirable in order to correct for self-enhancing.  Indeed there was no difference between rating of most and least desirable for participants and others, thus negating the self-enhancing possibility; i.e., the results could not be explained by participants' trying to self-enhance by attributing more desirable possibilities vs. undesirable possibilities to themselves vs. others.

Experiment 3: Hypothesis: The undergraduate participants were asked to consider desirable  future, undesirable future, or both, and the free will hypotheses predicted that they would see more possibilities overall for themselves than others.  Participants did choose more possibilities for themselves in the future than those of a friend.  by contrast, they did not chose the conjunction more than the desirable option when judging a peer. They would seem to indicate that the future was malleable i.e. not predetermined but in some way under their control for themselves.

Experiment 4: Hypothesis: The participants would view their own behavior in comparison to others', as the outcome of intentions and desires rather personality, history, or circumstances.  In order to test this hypothesis participants were asked to predict their own and their peers behavior by  use of boxes representing the situation, personality, desires and intentions, and past behavior.  The size of the box was to represent the predictive weight.  The experimenters found that participants weighted  their own desires and intentions relatively more than their personality, past behavior,and the situational variable.  In addition, their desires and intentions were weighted more than others' desires and intentions. Furthermore, participants viewed personality as the strongest predictor in roommates.  The experimenters actually calculated the areas of the boxes in making these calculations.  All of these data are consistent with their hypothesis.

What does all this mean with regard to free will?  This research certainly is strong support for the notion that people perceive themselves as possessing more of the qualities that constitute free will than others around them.  Participants in the experiments viewed their past and future behaviors as less predictable than their peers (experiment 1); participants believed they had more open paths in their lives that others (experiment 2 and 3); and the differences between self and other are furthermore a reflection the participants desires and intentions and this accounts for future behavior (experiment 4).