Causes of the Bias
So our attributions are often egocentric. After a failure performance, we formulate different set of explanations than
we might after success. We know this. The next question, and the one that has been burning up the journal pages for
the last decade is, Why? What is the cause of this egocentric bias?
At least three perspectives can account for the impact of outcomes on attributions (Forsyth, 1980). First, a number of
researchers feel that these attributional asymmetries are self- serving: when people succeed they can increase their
confidence and sense of personal worth by attributing their performance to internal, personal, or dispositional factors.
In constrast, when people fail, they can avoid the esteem-damaging consequences of their performance by denying
responsibility for their performance.
Second, a logical, information processing explanation like that proposed by Feather (1969; Feather & Simon, 1971)
emphasizes the relationship between anticipated outcomes and actually performance. According to this approach, if
people' outcomes match their expectations--they expect to succeed and pass or expect failure and flunk--then they
tend to attribute their outcomes to stable, internal factors such as ability. If, however, their outcomes violate their
expectations, then they attribute their outcomes to unstable factors; for example, luck, mood, or a more difficult test.
However, as Miller and Ross (1975) note, most people usually expect to do well since the covariation between (1)
their own behavior and positive outcomes and (2) the environment and negative outcomes is attributionally salient.
Thus, individuals tend to see themselves as the cause of positive performances, since negative expectations are rare.
Although it is likely that in instances of extreme and repeated failure a specific negative expectation will overwhelm
the generalized positive one, Miller and Ross maintain that in most achievement situations success, and not failure, is
expected.
Bradley (1978) has added a third possible explanation that emphasizes the interpersonal implications of attributions.
Since people' performances are often public and the subject of considerable discussion, people (1) attribute poor
grades to external factors to avoid the embarrassment of academic failure and (2) attribute good grades to their own
personal effort or ability to management the impression of competence. From Bradley (1978, p. 63): attributions are
"mediated by a desire to maintain or gain a positive public image (e.g., a public public motive) rather than by a concern
for one's private image".
These three explanations of the success-internal/failure- external pattern are not incompatible. Forsyth, in 1980,
presented a functional model of attributions that argued they fulfull four basic functions: explanation, predictive,
self-serving, and interpersonal. When ego-involvement or need for achievement is high then attributions may be biased
by self-serving motivations. However, people may also need to understand the causes of their outcomes if they are
going to improve after a failure, or maintain a level of success in the future. Therefore, they formulate explanatory,
adaptative attributions that explain the outcome, and suggest behavioral strategies for improvement or maintenance.
Lastly, if people wish to project a public image of ability and competence, then they may wish to make certain that
their teachers and classmates do not blame them for their failure, but do credit them with their successes. When
attributions fulfill an interpersonal function, then people can explain "What rotten luck!" or "The test was too hard"
after failure, and "I'm glad I worked as hard as I did!" or "Good, fair test" after success. This functional view of
attribution thus suggests that, dependent upon the circumstances, all three processes can combine to determine
attributions after success and failure.
This functional approach explains some inconsistencies in the literature dealing with the causes of self-serving biases.
First, the dimensions that we found when measuring unitary attributions didn't correspond to theoretical dimensions.
Attributions are evaluatively tinged: luck is either good or bad; ability is either high or low; tests are either good or bad.
Second, both the strict information integration view and the self-serving view overlook the emotional intensity of
reactions after success and failure.
Third, neither model can explain why egocentrisms are so pervasive. We defend any failure, even on the most
ridiculous of tasks.
Fourth, if externalization after failure insulates the self, then we should be able to find evidence of some benefit to the
self among individuals who do display this tendency.