1972). Intrinsic motivation, such as the pleasure that actors experience when
they practice a role on their own, is assumed to stem from the inherently re-
warding properties of the task rather than from positive environmental out-
comes of task performance, such as the reactions of an audience. However,
which rewards are perceived as apparent and beyond an activity ultimately
reflect a personal judgment by each learner, and as a result, extrinsic motiva-
tion can be difficult to study and interpret (Bandura, 1986).
Deci (1972) found that extrinsic outcome rewards, such as money or food,
reduced students’ intrinsic motivation on a task when these rewards were no
longer present. Lepper and Greene (1978) suggested that these adverse
means–ends effects were due to cognitive attributions of causation. Extrinsic
outcomes led students to attribute causation to these overt rewards, and this
led to decrements in engagement in the task when the rewards were no longer
present. Research by Lepper, Greene, and Nisbett (1973) indicated that de-
clines in motivation were the result of a means–end cognitive representation
rather than a reinforcement event. Participants who expected rewards for en-
gaging in an activity (i.e., the end) displayed decreased task engagement when
the rewards were withdrawn whereas participants who received extrinsic re-
wards unexpectedly did not show declines in task engagement when the re-
wards were withdrawn.
Although intrinsic motivation theorists often emphasize the adverse ef-
fects of extrinsic outcomes on intrinsically motivated individuals, these nega-
tive means–ends effects are contextually limited in several important ways
(Lepper & Greene, 1978). First, they occurred with participants having high
but not low initial motivation on a task, and second, these studies dealt with
performance rather than learning situations in which the participants already
possessed the requisite skills to engage in the activities and experience the sat-
isfying qualities. Participants without competencies in a task area, such as
children initially learning to read, do not show adverse effects when extrinsic
rewards are given (Zimmerman, 1985). In fact, extrinsic rewards can convey
knowledge of high competence at an activity, and these performance-
contingent rewards have been found to sustain intrinsic motivation whereas
rewards for mere task-engagement (i.e., task-contingent) have been found to
undermine it (Karniol & Ross, 1977). Thus, information about the ends of a
task in the form of external rewards does not detract from people’s valuing of
the intrinsic properties of a task (or means) if those rewards imply high com-
petence in controlling the means. Attributing outcomes to one’s underlying
competence may also influence the key social cognitive motivational variable
of self-efficacy beliefs.
There is evidence that extrinsic rewards for increasing competence en-
hance not only children’s task choice but also their self-efficacy beliefs and
task interest ratings (Zimmerman, 1985). Elementary school students receiv-
ing performance-contingent (i.e., process) rewards played significantly longer
330 ZIMMERMAN AND SCHUNK