DRAFT; please do not cite or quote
Multiple goals, values, and optimal motivation
Judith M. Harackiewicz and Amanda M. Durik
University of Wisconsin-Madison
Why do some students become involved and interested in their studies and why do they continue in a particular academic discipline? Why do some athletes become engaged in their sport, persist at practice, and seek competition against others? Answering these questions requires that we consider the processes underlying intrinsic motivation, or the motivation to engage in an activity for the value inherent in doing it (Deci & Ryan, 1985). We have studied the factors that influence optimal motivation and believe that goals play an important role in shaping intrinsic motivation and performance. To study goals and motivation, we have examined the role of intrinsic factors such as self-set goals and personal values in promoting interest and performance in academic contexts over time. We have also examined the effects of extrinsic factors such as goal interventions and task characteristics on intrinsic motivation in laboratory studies. How do these intrinsic and extrinsic factors combine to influence performance and ongoing motivation?
Our work has been guided by Harackiewicz and Sansone’s (1991; Sansone & Harackiewicz, 1996) process model of intrinsic motivation. Harackiewicz and Sansone draw an important distinction between the goals that are suggested or implied externally and the goals that are actually adopted by an individual in a particular situation (the perceived goal; see Figure 1). Rather than assume a one-to-one correspondence between an assigned goal and a personal goal, we suggest that the goals an individual adopts in a given situation can have multiple determinants. These effects are represented as A paths in Figure 1. One type of determinant involves situational factors, such as an experimental manipulation in a laboratory setting or a particular characteristic of a task. A second important type of determinant involves individual factors, such as initial interest, personal values, or personality differences in achievement orientation.
In this model, we focus on two levels of goals: purpose and target goals. Purpose goals reflect the reason for engaging in a task and represent what an individual hopes to accomplish in a particular situation. Target goals, on the other hand, reflect more specific goals for how an individual might achieve their overarching purpose goal (Bandura, 1986; Locke & Latham, 1990). For example, athletes may set a target goal of practicing a particular move three times each week. This specific standard serves as a more proximal mechanism to help them achieve their higher level purpose, which might be to improve their skills.
To better understand when and why goals have particular effects, moderator and mediator variables have been incorporated into the model. First, we consider the possibility that goals may have different effects in different contexts and/or for different types of individuals. In other words, the effects of goals on intrinsic motivation can be moderated by personality and situational factors, such that the direct effect of goals on motivation (the B path in Figure 1) can vary as a function of individual factors and/or the situation. Second, Harackiewicz and Sansone (1991) argued that goals at different levels of specificity can also interact such that higher order goals moderate the effects of lower order goals. Specifically, they suggested that the direct effect of target goals can be moderated by purpose goals. In other words, the direct effect of a target goal on intrinsic motivation can vary as a function of higher order purpose goals.
The question of why goals have particular effects is addressed with mediator variables which allow us to examine the underlying process through which goals affect intrinsic motivation. When mediation is established, the direct effect of a predictor on an outcome (the B path of goals to intrinsic motivation) can be better understood through the predictor’s effect on process variables (the C paths), which in turn influence the outcome (the D path). In particular, four key mediators of goals are shown in Figure 1: competence valuation, task involvement, perceived task value, and perceived competence. All four of these processes can promote intrinsic motivation. Specifically, individuals are more likely to experience intrinsic interest in an activity to the extent that they value doing well in the activity (competence valuation), become absorbed in the activity while engaged in it (task involvement), find meaning or value in the task (perceived task value), or feel competent at the activity (perceived competence). Including moderator and mediator variables has provided a richer understanding of when goals are likely to enhance intrinsic motivation and why they have these effects.
The Role of Achievement Goals in Optimal Motivation
We believe that intrinsic motivation is critical in any achievement endeavor and that an important indicator of success is whether individuals develop interest in their sport or course material and continue to play their sport or pursue further learning (Maehr, 1976; Nicholls, 1979). Indeed, most of our own research has focused on factors that increase or undermine intrinsic motivation. However, our competitive culture often defines success in terms of performance, measured in terms of how well a person performs relative to others. In a sports context, the criteria for success are clear – players win or lose. In an educational context, the most obvious indicator of success is academic performance, or grades, which may be based on normative curves. Of course, these may not be the only indicators of successful performance, but normative comparisons seem to underlie most conceptions of excellence in achievement situations. We therefore adopt a multifaceted definition of success and examine both performance and interest as indicators of optimal motivation.
In our first studies, we addressed the question of optimal motivation using an achievement goal approach. Achievement goals reflect the purpose or reason for an individual’s achievement pursuits in a particular situation (Ames, 1992; Dweck & Leggett, 1988; Maehr, 1989), and theorists have identified two general types of achievement goals: mastery and performance goals. When pursuing mastery goals, an individual’s reason for engaging in an achievement activity is to develop competence at an activity. In contrast, when pursuing performance goals, an individual’s reason for engagement is to demonstrate competence relative to others. Dweck and Leggett (1988) argued that achievement goals create a framework for how individuals approach and experience achievement tasks, and that mastery goals are more likely to foster an adaptive pattern of motivation and performance goals a maladaptive pattern. For example, early research suggested that when students pursued mastery goals they selected more challenging tasks, persisted in the face of difficulty, and held more positive attitudes toward learning. Conversely, students pursuing performance goals chose easier tasks, and withdrew effort when difficulty was encountered (Ames & Archer, 1988; Elliott & Dweck, 1988; Meece, Blumenfeld, & Hoyle, 1988; Nolen, 1988). The hypothesis that mastery goals are adaptive and performance goals are maladaptive will be referred to as the mastery goal perspective because it implies that only mastery goals can have positive consequences and that performance goals will have deleterious consequences.
Although there is little debate about the positive effects of mastery goals, others disagree with the second component of the mastery goal perspective. More recent reviews of the achievement goal literature suggest that strong conclusions about the negative effects of performance goals may be premature (Harackiewicz, Barron, & Elliot, 1998; Hidi & Harackiewicz, 2001). Some theorists endorse a multiple goal perspective in which mastery and performance goals are both considered adaptive (Harackiewicz, Barron, Pintrich, Elliot & Thrash, 2002). In particular, they argue that performance goals can have positive effects because they orient individuals toward competence and can promote adaptive achievement behaviors in some situations and for certain individuals.
We will review experimental work that reveals some of the conditions under which performance goals promote intrinsic motivation, as well as correlational work that reveals a positive association between performance goals and academic performance. We will then demonstrate how our findings, whether experimental or correlational, support a multiple goals perspective rather than the mastery goal perspective. Finally, we will discuss research that begins to integrate goals and values. Personal values can influence the achievement goals an individual adopts in a particular situation, and goals can influence the value that individuals come to find in tasks. Moreover, task values may also play an important role in the development of interest.
Evidence for Positive Effects of Both Goals from the Laboratory
In a series of experimental studies, Harackiewicz and Elliot (1993, 1998; Elliot & Harackiewicz, 1994) examined the effects of achievement goals on intrinsic motivation. The experimental activity in each of these studies involved playing an enjoyable pinball game. Before the start of the session, a mastery or performance goal was suggested to participants, who were all college students. Their mastery purpose goal manipulation highlighted development and improvement of pinball skills and the performance purpose goal manipulation highlighted normative comparisons and demonstration of pinball ability. Participants played two games, and performance was experimentally controlled to insure that each participant achieved a similar level of overall performance. After the second game, intrinsic motivation was measured with both self-report and behavioral measures.
Moderator Effects. The results revealed no main effects of achievement goals. Instead, the effects of assigned goals were moderated by achievement motivation (Jackson, 1974). Specifically, individuals low in achievement motivation (LAMs) showed higher levels of intrinsic motivation when assigned mastery goals, whereas individuals high in achievement motivation (HAMs) showed more interest when assigned performance goals. In explaining this pattern, Harackiewicz and Elliot noted that HAMs characteristically enter activities with a desire to increase their competence (Atkinson, 1974; McClelland, 1961). Assigning a mastery goal may not add much to how they typically approach achievement situations. A performance goal, however, provides the additional challenge of outperforming others and demonstrating competence, and thus may make the game more exciting and interesting for HAMs (Tauer & Harackiewicz, 1999). In contrast, LAMs typically avoid normative comparisons and experience performance anxiety in achievement settings (Atkinson, 1974). Assigning a performance goal can undermine interest for LAMs, but a mastery goal may help them appreciate their development of competence in the activity, increasing their interest in the game. In sum, neither achievement goal proved optimal for all participants.
Harackiewicz and Sansone (1991) argued that congruence between an individual’s higher and lower order goals is another key determinant of intrinsic motivation. Goals are congruent (or match) when they orient an individual to the same end. This matching hypothesis is consistent with other theories that suggest behavior is optimally regulated when lower order standards facilitate the attainment of higher level standards (see also Carver, Lawrence, & Scheier, 1996). Thus, target goals that help an individual achieve their purpose goal should optimize intrinsic motivation. Harackiewicz and Elliot (1998) tested the matching hypothesis with mastery and performance goal manipulations in another pinball study. Participants were assigned a mastery or performance target goal in the context of either a performance or neutral purpose goal. In this study, the effects of target goals on intrinsic motivation were moderated by purpose goals. Performance target goals enhanced intrinsic motivation more than mastery target goals when participants were given a performance purpose goal, whereas mastery goals enhanced intrinsic motivation more than performance target goals in the neutral control condition.
The findings from these pinball studies clearly suggest that performance goals can promote interest above baseline levels for some people (i.e., HAMs) and in some situations (i.e., when they match higher order goals), and are more effective than mastery goals in these cases. These findings indicate that both mastery and performance goals can promote interest, and that performance goals are sometimes superior to mastery goals. Inclusion of important personality and situational moderator variables revealed the conditions under which externally assigned mastery and performance goals can each enhance interest (see also Senko & Harackiewicz, 2002). Moreover, these effects highlight the complexity involved in influencing motivation with externally assigned goals. A given intervention may promote intrinsic motivation or not, depending on the individual and the context. We now turn to mediation analysis to help understand the motivational processes that were triggered within individuals in response to these goal manipulations.
Mediator effects. These findings raise the question of why both types of goals can increase interest in an activity. To address this question, we consider another type of third variable: mediators. Mediators represent the more proximal mechanism underlying the relationship between a predictor and outcome. In our work, we have identified four variables as mediators of intrinsic motivation: competence valuation, task involvement, perceived task value, and perceived competence. As noted earlier, our model suggests that intrinsic motivation can develop from placing greater importance on one’s competence in an activity (i.e., competence valuation), from becoming absorbed while engaged in an activity (i.e., task involvement), from finding value in the activity (i.e., perceived task value), and from feeling competent in the activity (i.e., perceived competence). For example, students’ intrinsic motivation would depend on whether they valued the skills and knowledge that they were learning, whether they became absorbed in class activities, whether they perceived the task as meaningful or important, and whether they developed a sense of competence in the material. However, in the research described below we will focus on the role of competence valuation and task involvement as mediators. Although we consider all three mediators important, our research has concentrated on the two processes initiated earlier in the motivational process (competence valuation and task involvement), because we have found these processes most relevant to goal effects. Specifically, goals can make individuals care more about doing well, and they can promote involvement in activities. For example, Harackiewicz and Elliot (1998) found that competence valuation and task involvement mediated the direct effects of goals on intrinsic motivation. Performance target goals were especially effective in promoting competence valuation and task involvement in the performance purpose goal condition, whereas mastery target goals had similar positive effects in the neutral purpose goal condition. In turn, higher levels of competence valuation and task involvement during the game promoted intrinsic interest in pinball, and mediated the goal matching effect. These results revealed a process in which participants first became affectively committed to attaining competence and then cognitively involved in the pinball game, resulting in increased intrinsic motivation.
In sum, the experimental work by Harackiewicz and Elliot revealed that both achievement goals could promote intrinsic motivation. Specifically, positive mastery and performance goal effects depended on personality differences (e.g., whether an individual was characteristically high or low in achievement motivation) or on characteristics of the situation (e.g., the match with other goals in the situation). Furthermore, by examining the underlying motivational process, Harackiewicz and Elliot found that mastery and performance goals facilitated interest through the same key mechanisms (competence valuation and task involvement). What proved more critical than the type of goal pursued was whether the goal fostered competence valuation and task involvement. These experimental results reveal that mastery and performance goals can both initiate positive motivational processes. Thus, this initial experimental work led us to consider a multiple goal perspective in which both goals could have positive consequences.
However, a limitation of these and other experimental studies is that participants have typically been asked to work on some activity under a mastery or a performance goal. Such designs force us into “either-or” inferences that compare one goal to the other, and tell us nothing about the additional benefits or disadvantages of pursuing a performance goal in conjunction with a mastery goal. Barron and Harackiewicz (2001) sought to test the effects of multiple goals by assigning both mastery and performance goals to participants in an experimental study, creating a multiple goal manipulation, which was compared to a mastery goal-only condition and a performance goal-only condition.They devised a laboratory version of an academic activity to simulate a classroom learning experience. Students were taught new methods for solving math problems. These methods used simple strategies to add, subtract, multiply, and divide complex numbers mentally (as opposed to more traditional strategies of working out problems with paper and pencil). The self-report and behavioral measures of intrinsic motivation were the same as those used in the pinball studies.