Attribution theory is “a theory about how people make causal explanations” (Kelley, 1973, p. 107). It’s a theory of how people perceive and determine the causes of results. Although attribution theory “was not formulated as a theory of individual motivation….[it] can be employed in the study of self-perception and in the formulation of a theory of motivation” (Weiner, 1972, p. 310) Applied to the individual learner, attribution theory is a theory of how the learner determines the causes of and is affected by their own successes and failures.
The basic assumption of attribution theory is that man is motivated to understand the causal structure of his environment, to know why an event has occurred, and to what source the event can be ascribed. Although there are multiple theories that fall under this label (Kelley & Michela, 1980, p. 458), the one which seems to have been most commonly adopted and appears to be most widely known in education is Weiner’s attribution-based theory of motivation (as evidenced by coverage of attribution theory in texts such as: Bohlin et al., 2009, pp. 281-282; Eggen & Kauchak, 1999, pp. 412-413; O’Donnell et al., 2007, pp. 150-151; Ormrod, 2003, pp. 409-412; Sternberg & Williams, 2010, pp. 376-377; Woolfolk, 2010, pp. 388-390).
Weiner’s theory was initially conceived in terms of the expectancy-value framework, supplemented with the motivation component of his mentor, John Atkinson (motivation x expectancy x value, 1957). However, concerned with restrictions of the predictions afforded by Atkinson’s theory, and in the face of hypotheses not being confirmed by experimental studies due to the length of time required by experimental subjects to complete his evaluations (Weiner, 2010, p. 30), Weiner began looking for other predictors and “more economical experimental procedures” (p. 30). Around 1968 he “rediscovered attribution theory and Fritz Heider” (p. 30), and by 1971 had reconceptualized on a foundation which combined ideas from both Heider and Rotter:
Heider and Rotter did not cite one another, although both were concerned with the perceived causes of success and failure and their locus or location. Rotter acknowledged one internal and one external cause, respectively, skill (ability) and luck (chance), whereas Heider intuited three causes (ability, effort, and task difficulty). I combined these two lists and proposed four main perceived causes of achievement outcomes—ability, effort, task difficulty, and luck (see Weiner et al., 1971). Two of these are internal to the person (ability and effort) and two are external (task difficulty and luck). It had taken me about 3 years to reach this very simple formulation! (Weiner, 2010, p. 30)
With this more solid conception he also created a graphical representation that provided a guide for his thinking and served as the foundation for subsequent theory building (Weiner, 2010, p. 31). His 2 x 2 representation embodied (a) four determinants of behavioral outcomes (ability, effort, task difficulty, and luck), (b) their two causal dimensions (locus and stability), and (c) links of value to causal locus and of expectancy to causal stability. Weiner’s graphical representation can be found in (2010, p. 32).
Weiner’s model was, in part, an extension of Heider’s (1958) model of effective forces, which stated that the action outcome, x, is “dependent upon a combination of effective personal force and effective environmental force” (Heider, 1958, p. 58). In other words, whether or not a person can do something depends on both his “personal power” (p. 99) and “environmental difficulty” (p. 99). With this model, predictions about future success are made based on past attributions:
The action manifestations, together with other “raw material” become the data that allow us, in a kind of factor analysis, to assess the role of the factors contributing to can. We assess when we attribute action outcome mainly to the person, mainly to the environment, or to a combination of both. Only then do we understand. Only then are we able to predict future action, for even when relatively momentary factors make an action possible, by circumscribing these factors, one acknowledges the existence of the more invariant and reliable personal and environmental conditions. (Heider, 1958, p. 99)
In Weiner’s model, Heider’s internal-external distinction was taken as the first of three dimensions to which an action result might be attributed. Weiner initially called this dimension locus of control after the use of Rotter (1954, 1966, as cited in Weiner, 1974), but later changed it to locus of causality (1985, p. 552). Locus of causality refers to the cause of the result of an action as being located either within the person, or outside in the environment. To this first dimension, Weiner, Frieze, Kukla, Reed, Rest, and Rosenbaum (1971, as cited in Weiner, 1974, p. 107) then added a second dimension of stability, to account for internal causal factors such as effort and mood, which are perceived of as being more variable than others, “changing from moment to moment or period to period” (Weiner, 1985, p. 551). Weiner (1979, as cited in Weiner, 1985) later added a third dimension, controllability, to account for factors that are under volition or optional control of the individual. These three dimensions (internal versus external, stable versus unstable, and controllable versus uncontrollable), together with the possibility of success or failure of an action, result in sixteen possible attributions. For example, if a student received a good grade on a test, the student might conclude that the grade was a result of his innate intelligence. This attribution would be classified as a successful outcome due to an uncontrollable, internal, and stable factor of intelligence, reinforcing the student’s self-estimation that he is smart. Contrariwise, if the student received a bad grade on a test and concluded that the grade was a result of his innate intelligence, this would again attribute the failed outcome to an uncontrollable, internal, and stable factor of intelligence. However, in this case, it would reinforce a self-belief that the he is dumb.
Weiner’s theory was further revised until it reached its final conception, a graphical representation of which can be found in Weiner (2010, p. 34). Weiner illustrated the temporal sequence of the model with two examples:
Assume a student fails an important exam. The initial experience following failure is unhappiness. Then there is a search for causality. Presume this person failed in the past even though she studies hard, whereas others succeeded on this exam. This pattern of information gives rise to the belief that the current failure is due to lack of aptitude. Aptitude is an internal, stable, uncontrollable cause, so there is a lowering of self-esteem, low expectancy of future success, hopelessness and helplessness, and shame and humiliation. Low expectancy (hopelessness) accompanied by these negative affects promotes the decision to, for example, drop out of school.
Now imagine that another student fails the same exam. This person also initially experiences unhappiness. But she has been successful in the past and the night before the test she was partying rather than studying. Her current failure therefore is ascribed to insufficient effort. This internal, unstable, and controllable cause lowers personal regard but also gives rise to the maintenance of expectancy, hope, guilt, and regret, all of which are positive motivators. Hence, motivation increases and she tries harder in the future. As noted earlier, prior failure or nonreinforcement can have positive or negative future motivational consequences. (Weiner, 2010, p. 33)
As demonstrated from Weiner’s examples, the resulting affective responses lead to a sense of self-worth and self-efficacy. Independent theories for both of these concepts will be discussed next.
 In support of this statement Weiner cites the following idea expressed by Heider (Heider, 1958, p. 79):
We shall be concerned with the actions of another person, in particular with the basic constituents of an action sequence which lead us to know that another person is trying to do something, intends to do something, has the ability to do something, etc. The concepts also apply to one’s own actions, but our main emphasis will be on actions in interpersonal relations.
 Both Kelley (1973, p. 107) and Weiner (1972, p. 313; 1985, p. 551) recognize Heider as the founder of attribution theory.
 Weiner (1985, p. 554) also noted two other possible dimensions for the model, intentionality and globality, but did not add them to the model because of a philosophical problem in the case of the former, and lack of evidence in the case of the latter.