Attribution Theory

B Weiner, University of California, Los Angeles, Los Angeles, CA, USA

ã 2010 Elsevier Ltd. All rights reserved.

An amalgam of assumptions, beliefs, data, hypotheses, The Perceived Causes of Success and
and theories related to phenomenal causality fall under Failure
the rubric of attribution theory. Hence, attribution theory
refers to a field of inquiry rather than to a specific scien- Explanations may be based on reasons or causes, which
tific conception. Inasmuch as the focus of study concerns need to be distinguished, although this differentiation is at
inferences about the causes of events and outcomes, times murky and fraught with philosophical intricacies
including achievement-related success and failure, and (see Buss, 1978). If a student is asked why he or she
the consequences of these beliefs, the field of attribution enrolled in a particular course, he or she might call
is of central concern to educators. Indeed, a great deal of forth reasons such as: ‘‘I need it for my major’’ ‘‘I heard
life in the classroom can be examined from an attribution the teacher is great’’ ‘‘It meets at the perfect time;’’ and so
perspective. forth. These explanations or justifications make the choice
The origin of attribution thinking for psychology understandable and intelligible. In explaining everyday
was Fritz Heider and his seminal book: The Psychology action, people typically call forth current reasons, which
of Interpersonal Relationships (Heider, 1958). However, there are associated with incentives (costs and benefits) and
had been a traditional philosophic interest in the study of volitional choice (see Malle, 2004).
causal reasoning and even prior to Heider a number of Attribution theory, on the other hand, is centered on
empirical studies were concerned with perceived causality causes. Causes are invoked to explain outcomes or end
(e.g., Michotte, 1946). Heider’s contribution was relatively results, such as success and failure, rather than actions;
neglected prior to the influential attribution analyses they are antecedents instead of (in addition to) justifications
offered by Jones and Davis (1965) and Kelley (1967). In and can apply to intended or unintended outcomes and to
addition to these contributions, the book Attribution: Perceiving factors that may or may not be controllable. Failing in
the Causes of Behavior (Jones et al., 1972) was also responsible mathematics because of perceived poor aptitude is consid-
for ushering in nearly two decades (1970–90) of attri- ered here to be an attribution, or a causal antecedent, rather
bution dominance in the field of social psychology. than a reason, while enrolling in a class to be with a friend is
Attribution theory, often associated with naive psychol- a reason (incentive) and is not regarded a cause.
ogy, replaced in impact the noncommon sense theory of It must also be noted that not all outcomes elicit a search
cognitive dissonance, and subsequently was supplanted for understanding and causality, for this requires cognitive
within social psychology by a focus on unconscious pro- work. For example, if one succeeds at a simple puzzle in the
cesses and priming. During and following its ascendance, newspaper, it is unlikely that time is taken to answer the
attributional thinking penetrated other fields of study as question: ‘‘Why did I succeed?’’ On the other hand, unex-
well, including educational psychology. pected failure at an important task is particularly likely to
This article presents aspects of attribution thinking give rise to a search to identify the cause (see Gendolla and
that are of relevance to education, confining this discus- Kolle, 2001). Attribution analyses thus are subsumed within
sion to the perceived causes of success and failure in cognitive functionalism in that future successful actions
achievement-related contexts. In so doing, a great deal often depend on perceiving the causes of past failures.
of attribution-related literature is omitted, particularly What, then, are the causal ascriptions for success and
from cognitive psychology pertaining to causal reasoning, failure? First, it must be remembered that attribution
from clinical psychology related to coping and mental theory is concerned with phenomenal causality, rather
health, and from social psychology concerning interper- than seeking the true causes. Perceived causes vary as a
sonal attitudes and behavior. function of situational context. For example, the perceived
The article proceeds as follows. Initially, the perceived causes of success and failure at sports (e.g., strength, a
causes of success and failure and their underlying proper- windy day) differ from the causes of success and failure at
ties or characteristics are considered. This is followed by a math (e.g., math aptitude). Math outcomes may even elicit
presentation of some antecedents of causal beliefs, and different dominant causal beliefs (e.g., aptitude) than do
then a more extensive examination of their cognitive, performance results at a history quiz (memorization
emotional, and behavioral consequences. effort). Causal beliefs also vary between age groups,

558
 Attribution Theory 559

cultures, and depend on whether the causal target is the enduring. Effort also is typically considered unstable,
self or someone else. although recall attribution theory deals with phenomenal
The potential causes of achievement-related outcomes causality so that labeling another lazy, or industrious,
therefore are numerous, diverse, and often idiosyncratic – implies stability in effort expenditure.
one must be wary of generalizations. Nonetheless, there is a Yet, a fourth causal characteristic relates to the globality
set of causal beliefs that appears in many contexts. The of the cause, or the extent to which the cause generalizes
most common causes of success and failure are aptitude, across situations. For example, it could be contended that
ability (or a learned skill), immediate and long-term effort, perceived general intelligence exerts an influence over a
task characteristics (such as ease or difficulty), intrinsic wide array of academic outcomes, whereas perceptual
motivation, teacher characteristics (such as competence), reasoning ability has more limited application.
mood, and luck (see Weiner, 1985, 1986). Furthermore, All causes, then, have multiple properties and can be
within this delimited list, aptitude (which is presumed classified within a taxonomic system. Aptitude, for exam-
here to include ability) and effort dominate causal beliefs. ple, is internal to the actor, stable, uncontrollable, and
Hence, one succeeded because one is smart and/or tried often considered global, whereas effort also is internal to
hard, and failed due to being incompetent and/or not the actor, but likely believed to be unstable, controllable,
exerting effort. At the very heart of the contribution of and specific. In a similar manner, chance tends to be
attribution theory to education is a conceptual analysis perceived as external to the actor, unstable, uncontrolla-
of the distinction between ability and effort and the con- ble, and specific. These groupings are important because
trasting linkages of these two causal beliefs to other cogni- causes that differ qualitatively (e.g., aptitude and chance)
tions, emotions, and actions. may share some consequences but differ on others. For
example, failing in math because of perceived lack of
aptitude or due to bad luck both capture the belief that
Causal Properties ‘‘I could not do anything about it’’ and give rise to the
results of uncontrollable thinking. On the other hand,
Inasmuch as causes are diverse and numerous, there have these causes are linked to different expectations about
been attempts to identify their underlying properties or the likelihood of future success because ability is stable
characteristics. For example, although it is evident that whereas luck is unstable.
ability and effort are not the same, in what ways do they
differ? Progress from description to classification allows
Causal Antecedents
causes to be compared and contrasted quantitatively,
rather than merely indicating that they differ qualitatively.
Attributional accounts of the determinants of causal
Certainly three, perhaps four, causal dimensions have
beliefs have been derived from an analysis of reasoning
been isolated (see Weiner, 1985, 1986). One characteristic
processes. The concepts most invoked are covariation,
of causes is their locus, or location, within or outside of
conditional probabilities, causal rules such as necessary
the actor. This causal property is most associated with
and sufficient causality, and the like (see Kelley, 1967).
Julian Rotter (1966), the originator of the concept referred
Discussions also often consider prescriptive or correct
to as locus of control. The locus label grew from research
reasoning, which is contrasted to the flawed conclusions
on skill-versus chance-determined tasks and is most asso-
reached by the person on the street. This body of knowl-
ciated with a scale used to classify individuals according
edge, related to epistemology, has applicability in educa-
to their beliefs regarding personal causality. Aptitude and
tional settings but is not discussed in this context.
effort are similar in locus, both being internal to the actor
More pertinent here is a consideration of some psy-
(although aptitude is likely to be regarded as more inter-
chologically based antecedents that give rise to causal
nal than effort), and differ from causes such as chance or
beliefs, particularly hedonic concerns. In addition, causal
task ease, which are regarded as located in the environ-
determinants that involve communications from teachers
ment, or external to the actor.
to their pupils also are of central importance in educa-
A second property on which causes can be compared
tional contexts. Finally, in the following section I consider
and contrasted is controllability. Aptitude, which is inter-
impression management techniques that attempt to alter
nal to the person, is not subject to volitional control and
the causal beliefs of others.
change, whereas effort expenditure is considered subject
to personal control – it could be otherwise.
Biases in Causal Reasoning
Endurance or stability is the third known property of
causes. Some causes, such as aptitude, are viewed as stable It has been repeatedly documented that individuals overes-
over time, whereas others, such as chance, are not timate themselves on virtually all positive characteristics.
560 Social and Emotional Aspects of Learning

Just as 75% of the mothers believe that their children are in not your fault and nothing could have been done.’’ Thus,
the upper 25% of the class, individuals regard themselves as a positive or pro-social emotional expression could have
smarter, more moral, better drivers, having more common negative behavioral consequences (see Graham, 1990).
sense, and so on, than others and compared to what they In a similar manner, praise for success at an easy task,
really are (see Dunning et al., 2004). A similar conclusion can and the absence of reprimand for failure at such tasks, are
be reached regarding causal reasoning: individuals relatively cues that one has low ability. Furthermore, uncalled-for
believe that they caused their own success, whereas failure help may be a low-ability signal from a teacher. In sum, a
was due to outside forces. This phenomenon has been given variety of causal information regarding ability and effort
various labels, including the self-serving attribution bias, has the teacher as its source.
ego-enhancement, and most prominently, the hedonic
bias. It is to be expected, then, that children and their
parents blame the teacher or the school for failure, while Impression Management Techniques
the teacher holds the child and/or his or her parents
It is intuitively evident that individuals do not want
responsible. Disagreement and conflict regarding the causes
others to blame them for failure or untoward actions. To
of success and failure are normative.
manipulate the causal beliefs of others, disparate im-
Other biases have been hypothesized and have been
pression management techniques are used. Four impression
very popular topics for attribution research. One consid-
management strategies have been identified: denial, excuse,
ered determinant of causal beliefs is labeled the actor–
justification, and confession. In denial, the actor does not
observer perspective. It has been suggested that actors
acknowledge a particular outcome (e.g., ‘‘I did not fail the
attribute their behavior to the situation, whereas obser-
test’’; ‘‘I was not smoking’’). This is a rather primitive
vers ascribe an action to a characteristic of the actor. For
technique, often used by younger children, for it can be
example, if I hit someone it is because that other person
readily disconfirmed if untrue.
provoked me (situational causality), whereas when some-
A more common strategy is to provide an excuse to
one else acts with aggression it is because that person is
change causal beliefs (ex ¼ from; cuse ¼ cause). The actor
hostile. Associated with this discrepancy is a dispositional
attempts to shift the perceived controllable cause (e.g., lack
bias, or the fundamental attribution error, which is the
of effort) to an uncontrollable one (e.g., the bus was late).
tendency to underestimate the situational influence on
In so doing, it is anticipated that blame and personal
the behavior of others and overestimate the importance
responsibility will be averted. Justification, a third identi-
of traits or perceived characteristics of the person.
fied impression management strategy, has a similar goal,
These presumed tendencies, however, are now being
although in this case allegiance to a higher moral goal is
called into question (see Malle, 2006). One factor under-
invoked (‘‘I did not study because I had to take my father to
mining these hypotheses is the established hedonic bias.
the hospital’’). Finally, confession of a misdeed is possible.
If individuals tend to take personal credit for success, then
In so doing, the wrongdoer acknowledges a bad behavior;
positive behaviors will not be ascribed by the actor to the
however, confessing puts forth the impression that he or
situation. On the other hand, in situations of failure, both
she is a good person who will not engage in the act in the
the hedonic bias and an actor–observer discrepancy go
future (‘‘I did not study for the test; I am sorry and it will
hand in hand, with over-attribution by the actor to the
not happen again’’).
situation, relative to an observer.
Teachers and their pupils at times disagree on the
acceptableness of various excuses and justifications. For
Teacher Communications example, pupils believe that their alarm failing to ring is a
reasonable cause for missing an appointment, whereas tea-
Teachers may inadvertently communicate causal infor-
chers do not regard this as acceptable to promote forgive-
mation to their pupils. This information can be conveyed
ness and lack of anger (see review in Weiner, 1995). It also is
through expressed emotions, praise and blame, and help
the case that some apparent excuses are truthful. However,
giving.
an observer cannot conclude with any accuracy whether a
As will be elaborated later, emotions are linked with
student in fact missed the test because the bus was late, or
causal beliefs. Anger following the failure of another tends
had the flu, as opposed to being negligent.
to be elicited by controllable causes, such as lack of effort,
whereas uncontrollable causality, such as lack of aptitude,
elicits sympathy. Compare, for example, parental reactions
to the failure of their child due to lack of studying as Causal Consequences
opposed to being caused by a mental handicap. Teachers
who express anger following failure convey to the pupils Of particular relevance to educators are the consequences
that the failure was their fault and is changeable, whereas of causal thinking on the expectations, emotions, and
heightened sympathy, and particularly pity, conveys ‘‘it is behavior of pupils. These are addressed in turn.
 Attribution Theory 561

Expectancy of Success beliefs. Hence, ascriptions of success to ability or effort

both promote pride and positive views of the self, whereas
Motivation theorists have established that behavior is in
attribution to luck, or to help from others, does not
part guided by future expectancy of success. Regardless of
contribute to these self-directed affects. Indeed, attribu-
the intensity of a need, if the expectancy of goal attain-
tion of success to others fosters gratitude (if the help
ment is low, then instrumental action directed toward that
provided by the other was volitional). One consequence
goal will not be undertaken. For example, regardless of
of the simultaneous influences of the hedonic bias as well
how hungry one is, a restaurant known to be closed will
as the desire to manage the thoughts of others is that
not be a target of choice.
public expressions of gratitude exceed private beliefs
To a great extent, expectancy of future success in
about the deservedness of this expression.
achievement contexts is determined by causal beliefs
Perceived causal controllability, the third property of
about the cause of prior failure. Stable or enduring causes
causes, has a complex relation to feeling states, with the
give rise to the conviction that the future will be no
linked affects dependent on the target of the emotion.
different than the past. Hence, attributions of failure to
Ascription of failure to a controllable cause such as lack
lack of aptitude, an uncaring school environment, an
of effort, given a desire to reach the goal, elicits guilt and
absence of support, and so on promote expectancy of
regret. On the other hand, lack of aptitude as the cause of
future failure. This lowers motivation and results in a
failure promotes humiliation and shame, which also
tendency to leave that setting. On the other hand, ascrip-
involve social comparison with competent others.
tion to changing factors provides hope that the future may
These same attributions for the behavior of others,
differ from the past. Attribution to lack of effort, or to a
however, elicit a different set of emotions. If another failed
teacher who is leaving, promotes positive expectations of
due to controllable causes, then anger tends to be elicited.
the future inasmuch as the perceived causes of the prior
On the other hand, nonattainment of a goal due to uncon-
failures are unstable or changing.
trollable causes tends to promote sympathy. Self- and
In an attribution-based change program designed to
other-related emotions are entangled within a complex
keep first-year college students from dropping out, it was
interconnected system – communicated anger, if accepted,
merely communicated that grades increase the longer the
implies a controllable cause, which should promote guilt,
student remains in school (Wilson and Linville, 1982).
whereas communicated pity implicates uncontrollable cau-
That is, over time, there is grade inflation. There was no
sality and promotes feelings of humiliation.
attempt to change the study habits, attendance, note tak-
A number of other achievement-related affects also
ing, or any other instrumental actions of the pupils. The
depend on causal thinking. Envy, for example, tends to
data suggest the effectiveness of this program in improv-
be elicited by internal and uncontrollable causality for the
ing school performance and decreasing dropout rates,
success of others. One is envious of another’s intellectual
presumably because expectation of success had increased.
capabilities, or beauty, but not of their hard work, since
one can also work hard. Admiration, on the other hand,
Emotions
can be evoked by the effort as well as the ability of
The appraisal approach to the understanding of emotion successful others.
is based on the simple assumption that feelings are deter- Affect-related personality inferences also are linked to
mined by thoughts, that is, thoughts are necessary and perceptions of the causal role of effort and ability. For exam-
sufficient determinants of affective states. For example, ple, if the success of another is perceived as due to ability,
negative self-related beliefs such as ‘‘I am a bad person’’ or while the other communicates effort or external factors as the
‘‘I am incompetent’’ are sufficient to promote negative cause, then modesty is inferred. Modesty is a desirable infer-
feeling states and decrease personal esteem. ence associated with liking. On the other hand, if ability is
Causal beliefs have far-reaching emotional conse- conveyed as the cause of success, then arrogance is inferred,
quences, both in regard to the self and feelings about even if it is the true cause. If Einstein says ‘‘I am an Einstein,’’
others. To understand the emotional lives of school chil- he nonetheless will be regarded as arrogant, which is an
dren, it is essential to invoke causal ascriptions. Here, I unfavorable trait. Thus, school-based competence has some
present an overview of the implications of causal thinking negative consequences in that others may be envious and
for some feeling states (see Hareli and Weiner, 2002). perceive the capable person as arrogant. This is perhaps
It has already been indicated that hope for the future responsible for pejorative labels, such as nerds and geeks,
and, similarly, the opposing feeling of hopelessness are which convey negative attitudes toward competence.
guided by the perceived stability of a prior cause of
failure. The other causal properties also have affective
Behavior
implications. Attribution of success to the self gives rise
to pride in accomplishment and increments in self- Expectancy of future success, along with affective states,
esteem, whereas this is not the case for external causal guides future action. That is, goal anticipations and
562 Social and Emotional Aspects of Learning

feelings bridge the gap between causal thinking and causes into fundamental properties was introduced. These
behavior. For example, assume that a student failed in underlying characteristics were shown to influence
an exam and ascribes that failure to lack of ability. Inas- expectancy of success, emotional reactions, and personal-
much as ability is a stable cause, expectancy of future ity inferences. Of special note is the array of affects linked
success is low; and being an internal and uncontrollable to causal thinking: admiration, anger, envy, guilt, grati-
cause, self-esteem decreases while shame and humilia- tude, hope, hopelessness, pity, pride, regret, self-esteem,
tion are experienced. Low expectancy of success, low shame, and sympathy were mentioned, and this is not the
self-esteem, and humiliation are motivational inhibitors full list of causally linked feelings.
that contribute to a decision to leave the setting. Expectancy and affect, in turn, are two key mediators
On the other hand, lack of effort is an unstable, con- of achievement-related behavior. Thus, increased school
trollable cause, so that attribution to this factor maintains motivation and decreased school dropouts can be affected
hope and positive anticipations, while arousing guilt, by alteration of perceived causality. In sum, the study of
which is a motivator to make amends for the past. Thus, education processes is intimately linked with attributional
an attribution of failure to lack of effort tends to be analyses.
adaptive and functional.
Along with the attributions, emotions, and behaviors of
See also: Cognition and Emotion; Mathematics Learning;
actors, this same sequence also describes observers of the
Motivating Students in Classrooms; Volitional Control of
actor. An observer making attributions of failure of another Learning.
to lack of ability tends to experience sympathy, which
promotes help-giving. On the other hand, ascriptions to
lack of effort evoke anger and an antisocial reaction (see
Weiner, 1995, 2006). Bibliography
As the attributions of ability versus effort are linked
with disparate reactions, attribution-change programs Buss, A. R. (1978). Causes and reasons in attribution theory:
have focused on having students make insufficient effort A conceptual critique. Journal of Personality and Social Psychology
rather than low-ability ascriptions in situations of failure 36, 1311–1321.
Dunning, D., Heath, C., and Suls, J. M. (2004). Flawed self-assessment:
(see Perry et al., 1993). These programs vary but often Implications for health, education, and the workplace. Psychological
provide models in film clips portraying adaptive attribu- Science in the Public Interest 5, 69–106.
tion patterns. The attribution interventions have proved Gendolla, G. H. E. and Koller, M. (2001). Surprise and causal search:
How are they affected by outcome valence and importance?
successful and provide promising tools to improve school- Motivation and Emotion 25, 327–349.
related performance. Graham, S. (1990). Communicating low ability in the classroom:
Bad things good teachers sometimes do. In Graham, S. and
Folkes, V. S. (eds.) Attribution Theory: Applications to Achievement,
Mental Health, and Interpersonal Conflict, pp 17–36. Hillsdale, NJ:
Some Concluding Comments Erlbaum.
Hareli, S. and Weiner, B. (2002). Social emotions and personality
inferences: A scaffold for a new direction in the study of achievement
Attribution theory was developed and nurtured by social motivation. Educational Psychologist 37, 183–193.
psychologists. However, the centrality of thoughts about Heider, F. (1958). The Psychology of Interpersonal Relations. New York:
causality in everyday life transported this body of knowl- Wiley.
Jones, E. E. and Davis, K. E. (1965). From acts to dispositions: The
edge to other fields of study. For education, attributional attribution process in person perception. In Berkowitz, L. (ed.)
analyses provided a ready fit inasmuch as parents, pupils, Advances in Experimental Social Psychology, vol. 2, pp 219–266.
and teachers focus so much attention on evaluation and New York: Academic Press.
Jones, E. E., Kanouse, D. E., Kelley, H. H., et al. (eds.) (1972).
the causes of good and poor performance. Attribution: Perceiving the Causes of Behavior. Morristown, NJ:
Two broad questions emerge from this approach: How General Learning Press.
does one know or determine one’s level of ability, the Kelley, H. H. (1967). Attribution theory in social psychology. In Levine, D.
(ed.) Nebraska Symposium on Motivation, vol. 15, pp 192–240.
adequacy of effort expenditure, and the like? Thus, exis- Lincoln, NE: University of Nebraska Press.
tential questions, such as ‘‘Who am I?’’ (as well as, ‘‘Who Malle, B. F. (2004). How the Mind Explains Behavior: Folk Explanations,
are you?’’), are addressed. In this article, three inferential Meaning, and Social Interaction. Cambridge, MA: MIT Press.
Malle, F. F. (2006). The actor–observer asymmetry in attribution:
sources were pointed out: biases that enhance the self, A (surprising) meta analysis. Psychological Bulletin 132, 895–919.
communications from teachers, and impression manage- Michotte, A. (1946). The Perception of Causality. New York: Basic
ment techniques used by others. Books.
Perry, R. P., Hechter, F. J., Menec, V. H., and Weinberg, L. E. (1993).
The second extended question asked is: ‘‘So what? Enhancing achievement motivation and performance in college
What difference does it make if success or failure was students: An attributional retraining perspective. Research in Higher
caused by low ability, as opposed to lack of effort?’’ Education 34, 687–723.
Rotter, J. B. (1966). Generalized expectancies for internal versus
These questions point out the functional and motivational external control of reinforcement. Psychological Monograph 80(1),
significance of causal thinking. Here, the classification of 1–28.
 Attribution Theory 563

Weiner, B. (1985). An attributional theory of achievement-related Weiner, B. (2006). Social Motivation, Justice, and the Moral Emotions.
emotion and motivation. Psychological Review 29, 548–573. Mahwah, NJ: Erlbaum.
Weiner, B. (1986). An Attributional Theory of Motivation and Emotion. Wilson, T. D. and Linville, P. W. (1982). Improving the academic
New York: Springer. performance of college freshmen: Attribution theory revisited. Journal
Weiner, B. (1995). Judgments of Responsibility: A Foundation for a of Personality and Social Psychology 42, 367–376.
Theory of Social Conduct. New York: Guilford.