The effect of monetary incentives on accounting student motivation.

Campbell, Steven V. ; Niles, Marcia S.

ABSTRACT

We investigate the effects of monetary incentives on accounting student motivation. Using a within-persons decision modeling approach and Russian accounting student participants to investigate potential motivators, we find students with monetary incentives placed significantly less emphasis on their overall grade-point average and significantly more emphasis on esteem in the eyes of classmates. Our results suggest monetary incentives do not undermine personal satisfaction, an intrinsic motivator. Finally, in the context of improving course performance, we find students with monetary incentives tend to value the attractiveness of academic success over the expectancy of success in making their effort-level decisions, whereas students without monetary incentives tend to value the expectancy of academic success over the attractiveness of success in making their effort-level decisions. These findings support the use of Vroom's (1964) expectancy theory as a conceptual framework for understanding accounting student motivation in cross-cultural settings.

INTRODUCTION

Incentives, the goal objects we desire to attain, figure prominently in several theories of motivation (e.g., Atkinson, 1964; Lewin, 1935; Rotter, Chance, and Phares, 1972; Vroom, 1964). In contrast to intrinsic incentives, extrinsic incentives involve, "the motivation to work primarily in response to something apart from the work itself, such as rewards or recognition or the dictates of other people" (Amabile, Hill, Hennessey, and Tighe 1995, 950). Most accounting educators readily acknowledge the importance of motivating their students, but disagreement exists concerning the use of extrinsic incentives for this purpose. Behavioral researchers often argue extrinsic incentives stimulate motivation and enhance academic achievement by making the learning objective more attractive. Cognitive researchers on the other hand generally contend extrinsic incentives undermine intrinsic motivation (an interest in learning for its own sake), decrease academic performance, and encourage a dependence on the acceptance, reinforcement, and approval of others (Bower, 1994).

In this paper we focus on the relation between monetary incentives and accounting student motivation. Specifically, in the context of improving course performance, we first examine the link between a performance-based monetary incentive and the attractiveness (valence) of academic success. Second, we examine the link between a performance-based monetary incentive and the amount of effort accounting students are willing to put forth to achieve academic success. Employing Vroom's (1964) expectancy theory and Stahl and Harrell's (1981, 1983) within-persons decision modeling method, our research design replicates earlier accounting student motivation studies by Harrell, Caldwell, and Doty (1985), Geiger and Cooper (1996), and Geiger, Cooper, Hussain, O'Connell, Power, Raghunandan, Rama, and Sanchez (1998). Our data are collected from 154 upper-level Russian accounting students. Monetary incentives are a prominent feature of student support for some, but not all, Russian university students and this differential treatment provided an element of control not available in other natural settings.

Our results support the applicability of expectancy theory and the within-persons decision modeling approach. Specifically, we find in making their effort-level decisions, Russian accounting students receiving a performance-based monetary incentive placed significantly less emphasis on improving their overall grade-point average and significantly more emphasis on increasing esteem in the eyes of their classmates, compared to accounting students not receiving a performance-based monetary incentive. We also find a performance-based monetary incentive did not undermine the personal satisfaction derived from superior academic performance. Finally, we find the students receiving performance-based monetary incentives tended to value the valence of academic success over expectancy of success in making their effort-level decisions, whereas students not receiving a performance-based monetary incentive tended to value the expectancy of academic success over the valence of success in making their effort-level decisions.

The next section provides some background information on higher education and the study of accounting in Russian universities. Section three develops the study's hypotheses and explains the within-persons research method. The results are discussed in section four and the fifth section concludes.

BACKGROUND

In recent years Russian higher education has experienced significant changes. During much of the Soviet period, the central government promoted a policy of full access to higher education and almost every secondary school graduate had his or her college education funded from the federal budget. Since the collapse of the Soviet Union, federal funding for higher education has been reduced and access to higher education is no longer universal. Most Russian universities have had to raise tuition and as a result the majority of students now come from high and middle income groups, which comprise about one-third of the Russian population (Smolentseva, 1999).

Russian university accounting programs are normally five year programs that students begin at age 17 or 18. Before enrolling in a Russian university, the typical Russian student must secure funding from one of three primary funding sources: the federal government, a private company, or parents. For purposes of this study, the important feature of federal government funding is a monthly cash stipend, the amount of which depends on the student's academic performance. At the Khabarovsk State Academy of Economics and Law, the setting for this study, the grades of federally funded students are reviewed at the end of each semester and if the student's academic performance fails to meet expectations, the student's monthly stipend is usually reduced or terminated. Generally, federally funded students with unsatisfactory grades receive no stipend, those with marginal grades receive a modest stipend, and those with high grades receive a more substantial stipend. These grade-based stipends are not available to privately funded students. Students with private company funding contract to work for a private company for a specified period of time after graduation in exchange for their college financing. Tax incentives are given to encourage private companies to enter into these student support contracts. Parental funding is less structured and only available to the wealthy.

Student funding from the federal budget is administered by the university and choosing the student recipients is an administrative decision. Students demonstrating superior academic ability at the secondary school level are likely to receive federal funding based on merit, but merit is not the sole criterion. Students whose parents are connected to the university or whose parents can benefit the university in some way are widely believed to receive preferential consideration in the selection process. Of the 154 student participants in this study, 47 were funded from the federal budget, 45 by private companies, 56 percent by parents, and six by other sources.

HYPOTHESIS DEVELOPMENT AND RESEARCH METHOD

Vroom's (1964) original formulation of expectancy theory consisted of two models, the valence model and the force model. In this study the valence model is expected to explain a Russian accounting student's perception of the attractiveness (valence) of academic success, defined in terms of receiving a high course grade. The force model is expected to explain the student's effort-level decision given the valence of the higher grade and the expected probability that an increased effort will result in the higher grade.

The Valence Model

The valence model captures the perceived attractiveness, or valence, of achieving a first-level outcome (academic success) by aggregating the valences of associated second-level outcomes (the potential motivators). In this study the valence of academic success, a first-level outcome, is determined by aggregating the valences of the following three second-level outcomes: (1) an improved overall GPA, (2) a strong feeling of personal satisfaction, and (3) increased esteem in the eyes of classmates. Thus:

[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] (1)

Where:

[V.sub.j] = the valence of the first-level outcome,

[V.sub.k] = the valence of the second-level outcome,

[I.sub.jk] = the perceived instrumentality, or belief, that [V.sub.j] will lead to [V.sub.k], and

n = the number of second-level outcomes.

The effects of extrinsic incentives on student motivation and learning have engendered a longstanding controversy in the education literature. This debate dates back to a study by Harlow, Harlow, and Meyer (1950) that found monkeys who had previously enjoyed solving puzzles would, after being rewarded with food for each puzzle solved, no longer solve puzzles when they were not given food. Conversely, monkeys that had not been rewarded with food continued to enjoy solving puzzles. Studies with humans since Harlow et al. (1950) have found similar results (e.g., Lepper, Greene, and Nisbett, 1973; Amabile, 1979; Deci and Ryan, 1985). A common explanation for these findings is that rewards cause people to lose interest in whatever it is they were rewarded for doing.

Our first hypothesis tests whether a monetary reward for achieving high grades is associated with a decrease in the influence of grades as a potential motivator. The studies cited in the preceding paragraph suggest a monetary reward for high grades will cause students to lose interest in grades as an independent motivating influence. If this is correct, improving overall grade-point-average (GPA) should be a less influential motivating influence for federally funded students, who receive a monetary reward, than it is for privately funded students who do not receive a monetary reward. Thus, our first hypothesis is:

H1: Improving overall GPA will be a less influential motivator for federally funded accounting students than for private company and parentally funded accounting students.

Motivation is not a single characteristic and in recent years numerous contrary terms have been proposed to describe the complex nature of the forces affecting student behavior. The oldest of these dichotomies is the distinction between intrinsic and extrinsic motivation. Many educators believe intrinsic motivation, an interest in the task for its own sake, is incompatible with extrinsic motivation in which the task is viewed as a prerequisite for obtaining something else (Deci and Ryan, 1985). They acknowledge rewards motivate students, but they contend this sort of motivation comes at the expense of personal satisfaction and other intrinsic values. Amabile (1979) and Harackiewicz and Elliot (1993) find extrinsic incentives undermine intrinsic motivation in college students. Furthermore, rather than helping students to develop their own criteria for successful learning, many educators and education researchers believe extrinsic rewards encourage a dependence on acceptance, reinforcement, and approval from others (Bower, 1994; Kohn, 1993).

Conversely, other theorists (e.g., Dweck, 1986 and McKeachie, 1961) have proposed college students have multiple learning goals and that mastery and performance goals can play complementary roles in motivating student learning. Under this view extrinsic incentives can either enhance or reduce interest in learning depending on how they are used. Cameron and Pierce (1994) find intrinsic motivation is not adversely affected by extrinsic rewards in many applications. They conclude educators need to abandon old beliefs about the negative effects of external rewards and embrace the idea of intrinsic and extrinsic motivational factors working together. Lin and McKeachie (1999) find medium levels of extrinsic motivation in combination with high levels of intrinsic motivation is more effective in facilitating college student learning than either low or high levels of extrinsic motivation.

In summary, the education literature concerning the relation between extrinsic incentives and intrinsic motivation is conflicting and controversial. In this study we address two issues related to the intrinsic and/or extrinsic motivation debate. First, we test whether a monetary incentive is associated with a reduction in the valence of personal satisfaction. Since only federally funded students receive a monetary reward for academic performance, we hypothesize:

H2: A strong feeling of personal satisfaction will be a less influential motivator for federally funded accounting students than for private company and parentally funded accounting students.

Second, we test whether a monetary incentive promotes an extrinsic orientation by fostering a dependence on approval from others. We hypothesize:

H3: Increased esteem in the eyes of classmates will be a more influential motivator for federally funded accounting students than for private company and parentally funded accounting students.

The Force Model

In the force model of Vroom's expectancy theory, Vroom hypothesized the motivational force influencing a person to act is a monotonically increasing function of the sum of the products of the valences of the second-level outcomes ([V.sub.j] in Equation 1) and the expectancy that the act will be followed by the attainment of these outcomes. Thus:

[F.sub.i] = ([E.sub.ij][V.sub.j]) (2)

Where:

[F.sub.i] = the motivational force to perform act i,

[E.sub.ij] = the expectancy that act i will result in outcome j, and

[V.sub.j] = the valence of outcome j.

The force model implies the motivational force acting upon a Russian accounting student to achieve academic success is explained by the sum of the valences of the second-level outcomes associated with academic success and the expectancy a particular effort-level will result in academic success. Success in this case is defined as earning a grade of "5" in an accounting course. (In Russian universities grades range from one to five with five being the highest mark.) Thus, the motivational force required to earn a grade of "5" is determined by the valence of earning a "5" and the expectancy that a particular level of effort will result in this outcome.

In this study the federally funded students have a second-level outcome not available to privately funded students, a performance-based monetary stipend. If the attractiveness of the other second-level outcomes associated with academic success are similar for federally and privately funded students, then additional valence of a monetary reward for federally funded students would cause the sum of valences for the second-level outcomes to be larger for federally funded students than for privately funded students. Thus, we hypothesize federally funded accounting students should be more influenced by the valence of academic success in their effort-level decisions than their privately funded counterparts:

H4: Regarding the relative influences of valence and expectancy in the force model, the federally funded accounting students will be more influenced by valence of academic success than private company and parentally funded students.

Within-Persons Decision Modeling

Decision modeling, as previously developed by Stahl and Harrell (1981, 1983) and employed in this study, involves a subject answering multiple decision making cases, each requiring separate decisions based on varying combinations of values for the second-level outcomes and the expectancy of success. Several prior studies have successfully used expectancy theory and the within-persons decision modeling approach to study accounting student motivation including Harrell et al. (1985), Geiger and Cooper (1996), and Geiger et al. (1998). This latter study demonstrated the appropriateness of the within-persons design for studying accounting student motivation in a cross-cultural context.

The within-persons approach avoids many of the methodological and measurement problems associated with an across-persons design (Kopelman 1977). By using each individual's decisions as operational measures of valence and expectancy an element of control is established. In this study we replicate the design used by Harrell et al. (1985) by testing the following three second-level outcomes: an improved overall GPA, a strong feeling of personal satisfaction, and increased esteem in the eyes of classmates. Each second-level outcome is manipulated at two levels, low (10 percent) and high (90 percent) and the expectancy of success is manipulated at three levels, low (10 percent), moderate (50 percent), and high (90 percent). This results in 24 decision cases, with each case presenting a unique mix of values for the three second-level outcomes and the expectancy of success. A sample case is presented in Exhibit 1. The decision cases were randomly ordered to prevent possible bias.

The participants in this study (n = 154) were third, fourth, and fifth year accounting students attending the Khabarovsk State Academy of Economics and Law, in the city of Khabarovsk in the Russian Far East. They completed the decision exercise during normal class time. Written and oral instructions were given at the time the decision exercise was administered. The students were told to assume they were at the mid-point of the semester and were currently earning a grade of "4" in an accounting course. The first decision (see Exhibit 1) asks the student to indicate the overall valence of increasing a grade of "4" to a grade of "5". The student's responses to the 24 decision cases are used to derive an individual regression model in which the student's valence decision is the response variable and the three second-level outcomes are explanatory variables.

The second decision in the decision exercise (see Exhibit 1) asks the student to indicate the level of effort he or she would be willing to exert to increase the grade, given their valence response in decision one and a stated probability of success. The outcome of this second decision indicates the motivational force acting on the student to increase the course grade. The student's responses to the 24 decision cases are used to derive a second individual regression model in which the student's effort-level decision is the response variable and the valence of success and the expectancy of success are the explanatory variables.

RESULTS

In this section we report the results of tests investigating the impact of monetary incentives on accounting student motivation. We first report the valence model results for our first three hypotheses and then we report the force model results for our fourth hypothesis. Table 1 shows the valence and force models were generally successful in predicting the valence and effort-level decisions of Russian accounting students. Of the 154 students who completed the decision exercise, 133 had significant individual valence and force models. Six of these students were funded from non-traditional sources and, since the nature of their funding arrangements were unknown, were not considered for further analysis. This yields a resulting sample of 127 student participants.

Valence Model Results

Table 2 presents a summary of significant individual valence model results for students in the three principal funding source categories. The mean [R.sup.2] (adj) statistics range from .68 to .72, indicating a uniform good fit. Improving overall GPA was the dominant motivator in all funding source categories and it was the highest standardized beta weight in 76 of the 127 individual regression models. Table 2 shows the mean standardized beta weights for increasing GPA is lowest for the government funded students, which is consistent with the first hypothesis. To determine whether the influence of GPA is significantly less for government funded students, we ran an unbalanced ANOVA in which the standardized beta weight for increasing GPA was the dependent variable and the three funding sources the independent class variable. This analysis indicated statistically significant differences (p=.014) across funding source categories. We also performed a second unbalanced ANOVA in which the private company funded students and the parentally funded students were pooled and compared to the government funded students. The results of this ANOVA indicated marginally significant differences (p=.062) in the influence of increasing GPA as a motivator. These results provide some support for the first hypothesis.

The second hypothesis predicts personal satisfaction will be a less influential motivator for government funded students than for privately funded students. The mean standardized beta weights reported in Table 2 indicate personal satisfaction was the second most influential motivator in all three funding source categories. Also, personal satisfaction was the highest standardized beta weight in 37 of the 127 individual regression models. As reported in Table 2, the mean standardized beta weights for personal satisfaction fall within the narrow range of .224 to .313. This suggests only minimal differences in the valence of personal satisfaction between the funding source categories.

To verify the lack of any significant treatment effect for personal satisfaction, we ran an unbalanced ANOVA in which the standardized beta weight for personal satisfaction was the dependent variable and funding source the independent class variable. This analysis indicated no statistically significant differences (p=.749) across the three funding source categories regarding the influence of personal satisfaction. We also performed a second unbalanced ANOVA in which the private company and parentally sponsored students were pooled. The results of this ANOVA also indicated no significant differences (p=.958) in the influence of personal satisfaction. These results do not support the second hypothesis.

The third hypothesis predicts increased esteem in the eyes of classmates will be a more influential motivator for government funded students than for privately funded students. Only 14 of the 127 individual regression models had increased classmate esteem as the highest standardized beta weight and the mean standardized beta weights reported in Table 2 also indicate this was the least influential motivator in all three funding source categories. However, consistent with our third hypothesis, the mean standardized beta weight for increased classmate esteem is higher for government funded students than for either private company or parentally funded students.

To determine whether the influence of classmate esteem is significantly higher for government funded students, we ran an unbalanced ANOVA in which the standardized beta weight for classmate esteem was the dependent variable and funding source the independent class variable. This analysis indicated marginally significant differences (p=.072) across the three funding source categories. We also performed a second unbalanced ANOVA in which the private company and parentally sponsored student groups were pooled and compared to the government funded students. The results of this ANOVA also indicated significant differences (p=.030) between government and privately funded students. These results provide some support for the third hypothesis.

Force Model Results

The high [R.sup.2] (adj) statistics presented in Table 3 indicate the force model was effective in predicting the students' effort-level decisions. Also the mean standardized beta weights reported in Table 3 for valence and expectancy, both within and across categories, suggest Russian accounting students consider both valence and expectancy important factors in their effort-level decisions. The private company and parentally funded students generally weighted the expectancy of success more heavily in their effort-level decisions, while the government sponsored students generally weighted the valence of success more heavily in their effort-level decisions. Slightly more than half of the students in the private and parentally funded categories indicated expectancy had a dominant influence in their effort-level decision, while slightly more than half of the students in the government funded category indicated valence had the dominant influence in their effort-level decisions.

To determine whether the influences of valence and expectancy differed significantly across funding source categories, we ran unbalanced ANOVAs in which the standardized beta weights for valence and expectancy were dependent variables and funding source the independent class variable. These analyses indicated no statistically significant differences across categories in the influence of valence (p=.160) or expectancy (p=.209); however, when private company and parentally funded students were pooled and compared to government funded students, we found marginally significant differences in the relative influences of valence (p=.078) and expectancy (p=.085). These results provide some rather weak evidence in support of the fourth hypothesis.

SUMMARY AND CONCLUSIONS

In this paper we use expectancy theory and a within-persons decision modeling approach to assess the influence of monetary incentives on accounting student motivation. Both components of expectancy theory, the valence model and the force model, were found to provide a useful conceptual framework for understanding the valence and effort-level decisions of Russian accounting students. Overall, our evidence suggests monetary incentives have subtle effects on student motivation.

We addressed four hypotheses. First, we found improving overall GPA was the dominant motivator for most Russian accounting students regardless of whether or not a monetary incentive was present. However, group-level differences in the emphasis placed on improving GPA were significant; students with monetary incentives were less influenced by the desire to increase their GPA than students without monetary incentives. Second, the education literature contains much debate on the relation between extrinsic incentives and intrinsic motivation. We found personal satisfaction, an intrinsic motivator, was influential for a considerable number of Russian accounting students regardless of whether or not a monetary incentive was present. In comparing students with a monetary incentive to students without a monetary incentive, we found no significant differences in the valence of personal satisfaction. The third second-level outcome examined, increased esteem in the eyes of classmates, was the least influential motivator regardless of whether or not a monetary incentive was present. However, consistent with out third hypothesis, group-level differences in the emphasis placed on classmate esteem were significant; students with monetary incentives were more influenced by classmate esteem than students without monetary incentives. Our fourth hypothesis concerned the force model of expectancy theory and in the context of improving course performance, we found students with a monetary incentive tended to value the valence of success over the expectancy of success in making their effort-level decisions, whereas students without a monetary incentive tended to value the expectancy of success over the valence of success in their effort level decisions.

One limitation of our study is the focus of expectancy theory on the individual decision maker. Our results provide educators with some useful general insights into accounting student motivation and the overall impact of monetary incentives on student motivation; however, motivation is an individual attribute and we observed considerable individual differences in our student participants' valence and effort-level decisions. This variation in individual results illustrates why expectancy theory is more properly applied on an individual level of analysis and why generalizations must be made with caution. Another limitation is the Russian student participants were not randomly sampled. Although we do not believe our sample selection procedures biased our results, the use of a convenience sample must be acknowledged. Finally, readers are cautioned that these results may not generalize to non-Russian students and non-accounting majors.

This research represents an initial application of expectancy theory and the within-person decision modeling approach to examine the relation between monetary incentives and accounting student motivation. The results suggest expectancy theory can provide accounting educators with a useful conceptual framework for understanding the effects of extrinsic incentives on student learning. It is hoped this initial effort will inspire accounting educators to use the natural controls provided by cross-cultural settings to examine the influence of other extrinsic incentives on accounting student motivation.

NOTE

Acknowledgement: The authors appreciate the assistance received from administrative officials and participating faculty at the Khabarovsk State Academy of Economics and Law.

REFERENCES

Amabile, T. (1979). Effects of external evaluation on artistic creativity. Journal of Personality and Social Psychology, 37, 221-223.

Amabile, T., Hill, K., Hennessey, B., & Tighe, E. (1995). The work preference inventory: assessing intrinsic and extrinsic motivational orientations. Journal of Personality and Social Psychology. 66, 950-967.

Atkinson, J. (1964). An introduction to motivation. Princeton, NJ: Van Nostrand.

Bower, B. (1994). Rewards surface as creativity enhancers. Science News. 145, 405-431.

Cameron, J., & Pierce, W. (1994). Reinforcement, reward, and intrinsic motivation: A meta-analysis. Review of Educational Research, 64, 363-423.

Deci, E., & Ryan, R. (1985). Intrinsic motivation and self-determination in human behavior. New York: Plenum.

Dweck, C. (1986). Motivational processes affecting learning. American Psychologist, 41, 1041-1048.

Geiger, M., & Cooper, E. (1996). Using expectancy theory to assess student motivation. Issues in Accounting Education. 11, 113-129.

Geiger, M., Cooper, E., Hussain, I., O'Connell, B., Power, J., Raghunandan, K., Rama, D., & Sanchez, G. (1998). Using expectancy theory to assess student motivation: an international replication. Issues in Accounting Education, 13, 139-156.

Harackiewicz, J., & Elliot, A. (1993). Achievement goals and intrinsic motivation. Journal of Personality and Social Psychology, 65, 904-915.

Harlow, H., Harlow, M., & Meyer, D. (1950). Learning motivated by a manipulation factor. Journal of Experimental Psychology, 40, 228-234.

Harrell, A., Caldwell, C., & Doty, E. (1985). Expectancy theory predictions of accounting students' motivation to achieve academic success. The Accounting Review, 60, 724-735.

Kohn, A. (1993). Punished by rewards: The trouble with gold stars, incentive plans, A's, praise, and other bribes. Boston: Houghton Mifflin.

Kopelman, R. (1977). Across-individual, within-individual and return of effort versions of expectancy theory. Decision Sciences, 41, 651-662.

Lepper, M., Greene, D., & Nisbett, R., (1973). Undermining children's intrinsic interest with extrinsic reward: A test of the "overjustification" hypothesis. Journal of Personality and Social Psychology, 28, 129-137.

Lewin, K. (1935). A dynamic theory of personality. New York: McGraw-Hill.

Lin, Y., & McKeachie, W. (1999, August). College student intrinsic and/or extrinsic motivation and learning. Paper presented at the 107th Annual Conference of the American Psychological Association, Boston, MA.

McKeaachie, W. (1961). Motivation, teaching methods, and college learning. In M.R. Jones (Ed.) Nebraska symposium on motivation, pp.111-142. Lincoln, NE: University of Nebraska Press.

Rotter, J., Chance, J., & Phares, E. (1972). Applications of a social learning theory of personality. New York: Holt, Rinehart & Winston.

Smolentsva, A. (1999). Current trends in Russian higher education. In International Higher Education, Center for International Higher Education, Boston, MA: Boston College Press.

Stahl, M., & Harrell, B. (1981). Modeling effort decisions with behavioral decision theory: Toward an individual differences model of expectancy theory. Organizational Behavior and Human Performance, 27, 303-325.

Stahl, M., & Harrell, B. (1983).Using decision modeling to measure second level valences in expectancy theory, Organizational Behavior and Human Performance, 32, 23-24.

Vroom, V. (1964). Work and Motivation. New York: Wiley.

Steven V. Campbell, University of Idaho

Marcia S. Niles, University of Idaho

Table 1: Sample Composition

Funding Number of Students with Significant
Source Students Valence and Force Models

Government 47 40 (83) (a)
Private 45 36 (80)
Parents 56 51 (91)
Other 6 6 (100)
Total 154 133 (86)

(a) Percent of students with significant valence and force models.

Table 2: Summary of Significant Valence Model Results by Funding Source

 Government Private Parents

[R.sup.2] (adj.) .72 .680 .680
GPA .487 .749 .619
SAT .271 .224 .313
EST .053 -.197 -.117

Correlations GPA SAT GPA SAT GPA SAT
EST -.26 -.22 -.30 -.34 -.54 -.31
GPA -.82 -.70 -.54

GPA = Mean standardized beta weight for grade point average. SAT = Mean
standardized beta weight for personal satisfaction. EST = Mean
standardized beta weight for esteem within the group.

Table 3: Summary of Significant Force Model Results by Funding Source

 Government Private Parents

[R.sup.2] (adj.) .93 .880 .920
VAL .524 .417 .457
EXP .499 .601 .577
Correlation -.990 -.970 -.970

VAL = Mean standardized beta weight for valence. EXP = Mean
standardized beta weight for expectancy

Exhibit 1. Sample Case from the Set of 24 Decision Cases

Model Elements If you receive a "5" in this course,
(not on the likelihood this will result in:
instrument):

Second-Level ...increased esteem in the eyes LOW (10%) (a)
Outcomes of your classmates is
([V.sub.k])

 ...a strong feeling of personal HIGH (90%)
 satisfaction is

 ...an improved grade point LOW (10%) (b)
 average is

Valence of First- DECISION A. With the factors and likelihoods shown
Level above in mind, indicate the attractiveness to you
Outcomes of receiving a "5" in this course.
([V.sub.j])
 -5 -4 -3 -2 -1 0 +1 +2 +3 +4 +5
 Very Very
 Unattractive Attractive

Expectancy of FURTHER INFORMATION. If you exert a great study
Success effort during the remainder of this semester, the
([E.sub.ij]) likelihood you will earn a "5" in this course is
 HIGH (90%).

Motivational DECISION B. With the attractiveness and likelihood
Force ([F.sub.i]) information above in mind, indicate the study
 effort you will exert for this course during the
 remainder of the semester.

 1 2 3 4 5 6 7 8 9 10 11
 Low Average Great
 Effort Effort Effort

 (a) Earning a "5" in this particular course is not
 likely to enhance your reputation in the eyes of
 your classmates.

 (b) It seems likely so much effort is required to
 earn a "5" in this course that doing so means your
 no improvement grades in other courses will
 suffer, resulting in your grade point average.