self-efficacy, perceptions of success, and intrinsic motivation for exercise

17
Self-Efficacy, Perceptions of Success, and Intrinsic Motivation for Exercise' EDWARD MCAULE? Universify of Illinois SUSAN WRAITH AND TERRY E. DUNCAN University of Oregon A number of investigators have attempted to identify the determinants of continued participation in exercise regimens. Within a social cognitive framework, the present study examined the relationships among self-efficacy cognitions, perceptions of suc- cess, and multidimensional intrinsic motivation in the exercise domain. Multivariate analyses of variance indicated that highly efficacious individuals were more intrin- sically motivated toward aerobic dance than were their less efficacious counterparts. However, multiple regression analyses revealed both perceptions of success and self- efficacy to account for significant variance in intrinsic motivation at both the compo- site and dimensional levels. Moreover, perceptions of success were considerably stronger predictors than were efficacy cognitions. The results are discussed in terms of the need for multiple measures of perceived success, the employment of longitudinal designs to determine the direction of causality among these variables, and possible implications for adherence issues. The U.S. Public Health Service (U.S. Department of Health and Human Services, 1979, 1980) has identified exercise and physical fitness as one of the 15 behavioral interventions likely to reduce death and disease, underscoring its importance as a major preventive health behavior. However, population estimates have indicated that 41 to 5 1 %of adults between the ages of 18 and 65 years of age currently remain sedentary (National Center for Health Statistics, 1980; Perrier Great Water of France, Inc., 1979). Of those Americans who do begin an exercise program, 50% are predicted to discontinue within the first six months (Dishman, 1982,1986; Morgan, 1977; Oldridge, 1982)and of those who are physically active, only 15% are estimated to expend the energy equivalent (1 500 kcal per week through exercise) suggested to have epidemio- I Preparation of this manuscript was facilitated by a grant to the first author from the National Institute on Aging (#R29 AG07907). The authors extend appreciation to Kimberly Poag, Jay Kimiecik, and two anonymous reviewers for their helpful comments on an earlier draft of this manuscript. 'Correspondence concerning this article should be sent to Dr. Edward McAuley, Department of Kinesiology, University of Illinois, Freer Hall, Urbana, 1L 61801. Journal of Applied Social Psychology, 1991,21 2, pp. 139-155. Copyright @ 1991 by V. H. Winston 8. Son, Inc. All rights reserved.

Upload: edward-mcauley

Post on 21-Jul-2016

245 views

Category:

Documents


12 download

TRANSCRIPT

Page 1: Self-Efficacy, Perceptions of Success, and Intrinsic Motivation for Exercise

Self-Efficacy, Perceptions of Success, and Intrinsic Motivation for Exercise'

EDWARD MCAULE? Universify of Illinois

SUSAN WRAITH AND TERRY E. DUNCAN University of Oregon

A number of investigators have attempted to identify the determinants of continued participation in exercise regimens. Within a social cognitive framework, the present study examined the relationships among self-efficacy cognitions, perceptions of suc- cess, and multidimensional intrinsic motivation in the exercise domain. Multivariate analyses of variance indicated that highly efficacious individuals were more intrin- sically motivated toward aerobic dance than were their less efficacious counterparts. However, multiple regression analyses revealed both perceptions of success and self- efficacy to account for significant variance in intrinsic motivation at both the compo- site and dimensional levels. Moreover, perceptions of success were considerably stronger predictors than were efficacy cognitions. The results are discussed in terms of the need for multiple measures of perceived success, the employment of longitudinal designs to determine the direction of causality among these variables, and possible implications for adherence issues.

The U.S. Public Health Service (U.S. Department of Health and Human Services, 1979, 1980) has identified exercise and physical fitness as one of the 15 behavioral interventions likely to reduce death and disease, underscoring its importance as a major preventive health behavior. However, population estimates have indicated that 41 to 5 1 %of adults between the ages of 18 and 65 years of age currently remain sedentary (National Center for Health Statistics, 1980; Perrier Great Water of France, Inc., 1979). Of those Americans who do begin an exercise program, 50% are predicted to discontinue within the first six months (Dishman, 1982,1986; Morgan, 1977; Oldridge, 1982) and of those who are physically active, only 15% are estimated to expend the energy equivalent (1 500 kcal per week through exercise) suggested to have epidemio-

I Preparation of this manuscript was facilitated by a grant to the first author from the National Institute on Aging (#R29 AG07907). The authors extend appreciation to Kimberly Poag, Jay Kimiecik, and two anonymous reviewers for their helpful comments on an earlier draft of this manuscript.

'Correspondence concerning this article should be sent to Dr. Edward McAuley, Department of Kinesiology, University of Illinois, Freer Hall, Urbana, 1L 61801.

Journal of Applied Social Psychology, 1991,21 2 , pp. 139-155. Copyright @ 1991 by V. H. Winston 8. Son, Inc. All rights reserved.

Page 2: Self-Efficacy, Perceptions of Success, and Intrinsic Motivation for Exercise

140 MCAULEY, WRAITH, AND DUNCAN

logic significance (Perrier Great Water of France, Inc., 1979, General Mills Inc., 1979).

Health benefits ascribed to physical activity have been reported to enhance both physiological and psychological well-being. From a physiolog- ical perspective, exercise has been demonstrated to reduce cardiovascular risk factors such as obesity, hypertension, and elevated blood lipids (Bon- nano & Lies, 1974; Cooper, 1968; Paffenbarger, 1986), as well as contribut- ing to the control of osteoporosis and Type I1 diabetes (Siscovick, Laporte, & Newman, 1985). The evidence linking exercise to psychological benefits is less compelling and certainly less consistent. However, a number of reports suggest that exercise is related to a reduction of symptoms of depression in healthy adults (National Institute of Mental Health, 1984) and possibly improves feelings of self-esteem, body image, and perceptions of control (Sachs, 1984). It should be noted, however, that no causal relationships have been established between exercise and these psychological states.

Numerous factors have been identified as possible influential variables in the adoption of and continued participation in exercise and physical activity (see Dishman, 1986; Dishman, Sallis, & Orenstein, 1985). For many individuals, it is unlikely that the activity in and of itself is sufficiently attractive initially to warrant continued participation. After all, aerobic activity entails time commitment, effort, persistence, and tolerance of likely discomfort and pain in the initial stages of adaptation. Incentives such as weight loss, improved physical attraction, and health concerns are possible catalysts in the initiation of exercise participation. For many individuals, it is likely that as conditioning and physical skills improve, intrinsic interest in the activity is enhanced. However, the cognitive processes that might mediate between gradual exercise success and intrinsic motivation have yet to be determined.

Vallerand and Reid ( 1984) have demonstrated that perceived success results in increased perceptions of competence that positively influence intrinsic motivation. Ryan (1982) and his colleagues (Deci & Ryan, 1985; Plant & Ryan, 1985; Ryan, Mimms, & Koestner, 1983) have proposed intrinsic interest or motivation to be a multidimensional rather than a uni- tary construct. Perceived competence, effort, interest-enjoyment, pressure- tension, and perceived choice are suggested to be first-order factors or dimensions representing a hierarchical or second-order unitary construct of intrinsic motivation. Thus, from this perspective it might be argued that gradual success in an exercise program influences one or more of the afore- mentioned dimensions, which in turn result in intrinsically motivated behav- ior. Vallerand and Reid (1984) have demonstrated that perceived success in a physical task can lead to increased perceptions of competence which posi- tively influences intrinsic motivation. Furthermore, McAuley, Duncan, and

Page 3: Self-Efficacy, Perceptions of Success, and Intrinsic Motivation for Exercise

SELF-EFFICACY AND INTRINSIC MOTIVATION 141

Tammen ( 1989) have recently reported evidence to support the hierarchical factor structure proposed by Ryan (1982).

From a social cognitive perspective, Bandura (1986) has contended that intrinsic interest is generated as a function of self-efficacy (Bandura, 1977, 1986) and self-evaluative cognitions such as perceived success. Such a propo- sition would imply that the continued interest of individuals engaged in exercise programs is cultivated by their self-beliefs in their abilities to exercise at given levels of frequency, intensity, and duration. Conversely, individuals who perceive themselves to be less competent in their ability to perform aerobic exercise routines at increasing intensity levels are likely to experience waning interest in such activity. Bandura (1986) has further proposed that there might exist a temporal lag between the acquisition of self-efficacy and the development of intrinsic interest in an activity. For many individuals, especially the sedentary, exercise may initially prove a less than pleasant experience even though some level of self-efficacy is perceived. However, with time, personal efficacy promotes mastery experiences leading to positive self-evaluation and, in turn, enhanced intrinsic interest. This self-evaluation mechanism can be determined in a number of ways. For example satisfac- tion/ dissatisfaction with goal attainments has been operationalized as such a mechanism (Bandura & Cervone, 1983). Similarly, tapping subjective percep- tions of performance success can be employed as a self-evaluative mechanism (McAuley, 1985; McAuley & Tammen, 1989; Spink & Roberts, 1980).

Although a number of investigations have implicated aspects of intrinsic motivation and self-efficacy to play a role in exercise behavior (e.g., Wankel, 1985; Dzewaltowski, 1989), little effort has been made to determine how these variables are related with respect to the process of exercise behavior. If self-efficacy and self-evaluative mechanisms contribute to increased interest in exercise activity, interventions designed to facilitate these factors in organ- ized programs might contribute to a reduction in recidivism. To this end, the purpose of the present investigation was twofold. First, we sought to test Bandura’s ( 1986) hypothesis that differing levels on self-efficacy should have differential effects on intrinsic motivation. Second, the degree to which per- ceptions of success and personal efficacy were related to intrinsic motivation for aerobic exercise activity was determined.

Method

Subjects

Participants enrolled in university aerobic dance classes were recruited as subjects for this study. The sample totaled 265 subjects and included 254 females and 11 males. Subjects were drawn from three levels of aerobic

Page 4: Self-Efficacy, Perceptions of Success, and Intrinsic Motivation for Exercise

142 MCAULEY, WRAITH, AND DUNCAN

exercise classes, beginners ( N = loo), intermediate ( N = 109), and advanced ( N = 56). The mean age of the sample was 20.73 years (SD = 3.98).

Measures

Intrinsic motivation. The Intrinsic Motivation Inventory (IMI; Ryan, 1982) was employed to assess overall intrinsic motivation and the underlying dimensions of interestenjoyment, perceived competence, effort, pressure- tension, and perceived choice. The IMI appears to have several unique psy- chometric qualities. Comprised of 27 items, the full set of items have rarely been used and the inclusion/exclusion of any one factor/dimension fails to affect adversely the remaining factors. Furthermore, shorter versions of each subscale have been employed with little adverse effect on internal consistency, consequently redundant items can be eliminated. The individual items due to their generic substance can also be easily modified to suit a wide variety of activities. The IMI has been employed by Ryan and his colleagues in such achievement tasks as writing, reading, and learning. Recent work by McAuley and his colleagues (McAuley et al., 1989; McAuley & Tammen, 1989) reports acceptable validity and reliability for a version of the IMI employed in competitive sport settings.

The version of the IMI utilized in the present investigation was comprised of 27 items suitably worded to reflect the situation of interest, aerobic dance, and was scored on a Likert scale from 1 (strongly disagree) to 7 (strongly agree). The scale included a minimum of four items for each of the five subscales and was considered of practical length for administration. Each subject obtained a score for each dimension and a total intrinsic motivation score, computed by summing the dimensional scores.

Self-efficacy. Percepts of personal efficacy with respect to exercise behavior were assessed by having the subjects indicate whether they were capable of exercising at their target heart rate (intensity of activity), for a given period of time (duration of activity), three times per week (frequency of activity). The scale was composed of five items beginning at I0 min and escalating in 10-min intervals to 50 min. An example of a scale item was “I believe that I can exercise at my target heart rate for 10 minutes without stopping three times per week.” If subjects responded positively in terms of being able to achieve a given item/level, they were asked to rate their confidence in this prediction on a scale of 10 (highly uncertain) to 100 (completely certain). Confidence scores were then summed and divided by the total number of items to attain a strength of self-efficacy score. Such procedures are in accord with Bandura’s ( 1977) general guidelines for assessing self-efficacy.

Perceptions of success. Subjects’ perceptions of success in the exercise class were assessed by one item scored on a 1-7 Likert scale indicating how success-

Page 5: Self-Efficacy, Perceptions of Success, and Intrinsic Motivation for Exercise

SELF-EFFICACY AND INTRINSIC MOTIVATION 143

ful they felt they had been in improving their generalphysical condition as a function of class participation. Although it can certainly be argued that there are a multitude of ways to assess different aspects of subjective success, we were concerned here only with perceptions of success regarding physical condition. Rationale for this stems from the focus of the activity, physical conditioning.

Procedures

Prior to the start of the 10-week-long classes in aerobic dance, the exercise leaders/instructors were approached in regard to the nature of the study and permission secured to recruit subjects from these classes. On the first session of the final week, subjects were informed that the general purpose of the study was to examine students’ opinions of aerobic exercise and then provided with written informed consent forms. The nine week period between the beginning of class and collection of data was considered sufficient to formulate cogni- tions regarding exercise efficacy (for the beginners) and intrinsic motivation for exercise. Upon reading and completing these forms, a short background questionnaire was administered, followed by the IMI, self-efficacy, and per- ceived success measures.

Results

The data were analyzed in two stages. First, the psychometric properties of the IMI in the exercise setting were examined by conducting hierarchical confirmatory factor analyses to determine the tenability of the hypothesized factor structure, comprising one higher order factor of intrinsic motivation with five underlying dimensions of interest-enjoyment, effort, pressure- tension, perceived competence, and choice. In the next stage the relationships among perceptions of success, efficacy, and intrinsic motivation were exam- ined. Multivariate analyses of variance were employed to determine which dimensions of intrinsic motivation maximally differentiated high and low efficacious individuals. Finally, hierarchical multiple regression analyses were conducted to discern the relative contributions of perceptions of exercise success and efficacy cognitions to overall intrinsic motivation as well as the individual dimensions.

Confirmatory Factor Analyses

Before conducting the confirmatory factor analyses, item-reduction tech- niques were utilized to determine whether all of the individual items contrib- uted sufficiently to the overall construct and underlying dimensions to war-

Page 6: Self-Efficacy, Perceptions of Success, and Intrinsic Motivation for Exercise

144 MCAULEY. WRAITH, AND DUNCAN

rant inclusion in subsequent analyses. These analyses resulted in the exclu- sion of six items with low inter-item correlations. Following scale reduction, the internal consistency of the five subscales was assessed via coefficient alpha (Cronbach, 1951). The final 21-item scale employed in the study is shown in Table 1. With the exception of the choice dimension (a = .54), the reliabilities were all quite acceptable with values ranging from .72 to .93. All alpha coefficients and interscale correlations are displayed in Table 2.

The LISREL VI (Joreskog & Sorbom, 1984) computer program was employed to verify the proposed hierarchical factor model. The hypothe- sized factor structure was composed of the 21 observed variables (items) loading on the five first-order (dimensions) factors underlying a single second-order factor of intrinsic motivation. In higher-order models, the co- variances among first-order factors are set to zero, and each of the first-order factors is hypothesized to load on a single higher-order factor. The standard- ized maximum likelihood factor loadings of the observed variables (items) on the latent variables (dimensions) for the first- and second-order models are shown in Table 3.

As can be seen, the loading of the items on their hypothesized dimensions are all statistically significant, as denoted by z-values greater than 1.96. However, the relationship between the choice dimension and overall intrin- sic motivation is largely explainable by one item (“I participate in aerobics because I have no other choice”). Coupled with the previously reported low internal consistency, these data suggest caution in any future use of this particular scale, Indeed, no data currently exist, of which these authors are aware, attesting to the psychometric properties of the perceived choice subscale.

In order to assess the fit of these data to the proposed factor structure a number of model fitting procedures were administered. The chi-square goodness-of-fit test, goodness-of-fit index, and root mean square residual (RMSR) statistics were generated as a function of the LISREL program. These are shown in Table 4. Additionally, the proposed factor structure was compared to a null or completely independent model and the goodness-of-fit statistic, coefficient delta (Bentler & Bonett, 1980), subsequently calculated. It is important to emphasize that even in the event of the hierarchical model being able to account for all of the covariation among the five first-order factors, the goodness-of-fit indices cannot be improved over those of the first-order model. However, if the fit indices for the hierarchical model approach those of the corresponding first-order model then the hierarchical model can be interpreted as being a more parsimonious fit to the data (Marsh, 1985). Although the difference in chi-square values between the first and second-order models is statistically significant, the relative differences between the other fit indices for the competing models is negligible, suggest-

Page 7: Self-Efficacy, Perceptions of Success, and Intrinsic Motivation for Exercise

SELF-EFFICACY AND INTRINSIC MOTIVATION 145

Table I

The Intrinsic Motivation Inventory with Items Worded to Reflect Aerobic Exercise

1. I enjoy participating in this aerobics class very much (I-E) 2. I think I am pretty good at aerobics (PC) 3. I put a lot of effort into this aerobics class (E-I) 4. I do not feel nervous at all while participating in aerobics (P-T) 5 . This aerobics class is fun to do (I-E) 6. I think I d o pretty well in this aerobics class, compared to other students

7. I haven't tried very hard to do well in this aerobics class (E-I) 8. I feel very tense while participating in this aerobics class (P-T) 9. I haven't really had a choice about participating in this aerobics class (C)

10. I think this aerobics class is boring (I-E) 1 I . I try very hard in this aerobics class (E-I) 12. I am very relaxed in doing the routines presented in this aerobics class (P-T) 13. I feel like I have to participate in this aerobics class (C) 14. This aerobics class does not hold my attention at all (I-E) 15. I would describe this aerobics class as very interesting (I-E) 16. I am pretty skilled at the level of aerobics presented in this class (PC) 17. I haven't put much energy into this aerobics class (E-I) 18. I feel pressured during participation in this aerobics class (P-T) 19. 1 think this aerobics class is quite enjoyable (I-E) 20. While participating in this aerobics class, 1 think about how much I enjoy

aerobics (I-E) 21. I participate in this aerobics class because I have no other choice (C)

(PC)

Legend: I-E = Interest-Enjoyment; PC = Perceived Competence; E-I = Effort- Importance; P-T = Pressure-Tension; C = Choice.

Table 2

Correlations and Reliabilities for the Five Intrinsic Motivation Dimensions

3. 4. 5. 2. 1.

I . Interest-Enjoyment (.92) 2. Competence .321 (.83) 3. Effort .618 .474 (37) 4. Pressure-Tension .419 .713 .445 (.72) 5. Choice .555 .260 .374 .333 (34)

Note. Reliabilities are displayed in parentheses, and correlations among the dimen- sions in the subdiagonal matrix.

Page 8: Self-Efficacy, Perceptions of Success, and Intrinsic Motivation for Exercise

146 MCAULEY, WRAITH, AND DUNCAN

Table 3

Parameter Estimates for the 21 Intrinsic Motivation Items for the First- and Second-Order Models

First-order Second-order

Effect r-value Effect t-value Coefficient Coefficient

Loadings on the latent variables

Enjoyment ( I ) lambda x l , I .863 lambda XS, I ,919 lambda X I O , .785 lambda xl4. .7 16 lambda XIS, I .818 lambda x19. ,916 lambda x20. ,561

lambda x2. 2 ,827 lambda x6, .664 lambda x16. ;! 366

lambda x3, 3 ,846 lambda x7. 3 .697 lambda X I ] . 3 5 5 lambda x17, ,794

lambda x4, ,538 lambda x8.4 ,669 lambda x12. .I ,749 lambda X I & ., .594

lambda x9, 5 S O 5 lambda XI) , .272 lambda x21, 5 .953

Competence (2)

Effort (3)

Pressure-Tension (4)

Choice ( 5 )

17.471 19.404 15.090 13.249 16.058 19.296 9.723

15.323 1 1.472 16.322

16.397 12.444 16.664 14.925

8.457 10.975 12.610 9.503

7.444 4.180

11.799

lambda y l , lambda y5, lambdaylo, I

lambdayl4, I

lamb day^^. I

lambdayl9, I lambda y 2 0 . I

lambda y2, lambda Y6, 2

lambda Y16. 2

lambda y3. lambda y7.3

larnbdayl~, 3

lambda Y17, 3

lambda y4, lambda YN, 4

lambda y12. 4

lambda Y I S , 4

lambda x9, 5

lambda x13, 5

lambda x21, 5

365 ,920 ,785 .714 316 .9 15 .563

,819 .673 370

346 .702 .852 .795

.518

.696

.725 ,610

.535

.282

.903 Intrinsic Motivation gamma I , I .755 gamma 2, I ,599 gamma 3, I .755 gamma 4. .665 gamma 5 . I ,609

21.652

17.517 14.734 18.975 25.167 10.343

-

13.352 11.190 -

14.833 11.876 14.949 -

-

6.999 7.101 6.577

5.826 3.822 -

-

7.289 8.321 5.854 7.462

Legend: Loadings detailed above depict strength of association between IMI items and the theorized dimensions. For example, in the second-order model lambda x6, 2

represents the loading of item #6 on the dimension of competence, whereas gamma 2, I

indicates the strength of the loading of the perceived competence dimension on intrinsic motivation.

Page 9: Self-Efficacy, Perceptions of Success, and Intrinsic Motivation for Exercise

SELF-EFFICACY AND INTRINSIC MOTIVATION 147

ing that the hierarchical model provides an adequate approximation of the data.

From the preceding analyses, it appears that the IMI employed in the exercise setting has reasonably good construct (factorial) validity. Although examination of the residual correlation matrix revealed some aberrant values, the relatively high value of coefficient delta (39) and the small magnitude of the RMSR in the first-order model indicates a tenable factor structure. It is entirely possible that the overall fit of the model could be improved substan- tially with item addition and modifications to theperceived choice dimension. Indeed, concern with the reliability and factor loadings of items on the choice dimension prompted us to adopt a conservative stance by omitting this dimension from subsequent analyses.

Preliminary Analyses

As the subject population was drawn from three levels of aerobic activity (beginner, intermediate, and advanced), preliminary analyses were conducted to determine whether this grouping differentially affected self-efficacy and intrinsic motivation. Certainly it would be expected that more advanced participants would display higher perceptions of efficacy. Should differences be present between groups along the dimensions of intrinsic motivation, this may serve to confound any relationships present between intrinsic motivation and self-efficacy.

A one-way analysis of variance with Scheffe post-hoc follow-up tests revealed a significant difference between group level on self-efficacy, F(2, 262) = 6 . 4 5 , ~ < .01. The beginning group had lower aerobic self-efficacy

Table 4

Comparison of the Fit Indices for the First- and Second-Order Models of 21 Item Intrinsic Motivation Inventory

Coefficient ~

First-order Second-order

Goodness-of-fit indices

XYdf 1 357.05 (179) 428.61 (184)

x2(df) null model 3250.38 (210) 3250.38 (210)

delta VO - f d / f ~ .89 .87

Adjusted goodness-of-fit .846 .824

Root mean square residual .057 ,075

Note. All chi-square values are significant (p < .05).

Page 10: Self-Efficacy, Perceptions of Success, and Intrinsic Motivation for Exercise

148 MCAULEY. WRAITH, AND DUNCAN

(M = 63.87, SD = 20.67) than the intermediate group (M = 67.56, SD = 32.01), which was less efficacious than the advanced group ( M = 77.03, SD= 22.71). The advanced group differed significantly from the other two groups, which were not significantly different from each other. The pattern of these results is in keeping with the tenets of self-efficacy theory (Bandura, 1977, 1986) and other studies that suggest more efficacious individuals persist longer in exercise activities (McAuley & Jacobson, in press) therefore becom- ing more experienced.

A one-way multivariate analysis of variance (MANOVA) between level of group along the dimensions of intrinsic motivation was nonsignificant, F(8, 522) = 0.249, p = ns. Thus, if any differences in intrinsic motivation among subjects exist they appear not to be directly related to level of aerobic activity but to some other construct.

Multivariate Analyses

Multivarite analyses of variance (MANOVA) were conducted to determine whether levels of exercise self-efficacy resulted in differential effects on in- trinsic motivation. Subjects were classified as being high or low on self- efficacy if they fell into the top or bottom thirds of overall self-efficacy score.3 The dimensions of effort-importance, enj oyment-interest, perceived compe- tence, and pressure-tension comprised the four dependent variables. Mean values and standard deviations are shown in Table 5 . The overall multivariate effect was significant, F = 6.24, p < .0001 (Wilks' criterion). Subsequent univariate analyses indicated that highly efficacious individuals perceived themselves to enjoy aerobic exercise more, put forth more effort, experience more pressure-tension, and feel more competent than their less efficacious counterparts. However, examination of the follow-up discriminant function coefficients indicated that perceptions of exercise competence (.68) maxi- mized the multivariate effect between the two efficacy groups. Such a relation- ship is perhaps to be expected, as perceived competence is clearly the dimen- sion of intrinsic motivation most closely aligned to efficacy cognitions.

Multiple Regression Analyses

A primary objective of the present study was to determine whether, as the social cognitive perspective (Bandura, 1986) proposes, self-evaluative (i.e., perceptions of success) and self-efficacy mechanisms influence intrinsic moti- vation for aerobic exercise. As intrinsic motivation is considered here to be

'Similar multivariate analyses were conducted on a median split of the data and revealed essentially the same results. Therefore, only the data for the extreme groups are reported.

Page 11: Self-Efficacy, Perceptions of Success, and Intrinsic Motivation for Exercise

SELF-EFFICACY AND INTRINSIC MOTIVATION 149

Table 5

Means and Standard Deviations for High and Low Efficacy Groups Along Dimensions of Intrinsic Motivation

Dependent variable High efficacy Low efficacy

Pressure-Tension 30.75 (3.35) 28.89 (4.75)

Effort-Importance 29.50 (5.02) 26.35 (6.33)

Perceived Competence 35.70 (4.50) 32.1 1 (5.56)

Interest-Enjoyment 41.58 (7.60) 37.56 (8.46) ~ ~ ~~

Note. All mean differences are significant @ < .01, df = I , 182). High efficacy, N = 97, low efficacy, N = 87.

comprised of several dimensions, each in itself important, a series of hier- archical regression analyses, employing the entire sample population, was conducted. In these analyses, overall intrinsic motivation and each of the individual dimensions were treated as separate dependent variables and self- efficacy and perceptions of success as the predictor variables. For each dependent variable two regression equations were calculated to allow each predictor variable to be entered last into the equation, thereby identifying the unique variance contributed to the dependent variable by that predictor. Once the unique variances were determined for each predictor, they were summed and subtracted from the total variance accounted for to provide an index of the variance shared by the two predictor variables.

Table 6 details the results of these analyses. As can be clearly seen, perceived success was the major predictor of overall intrinsic motivation accounting for 28.7% of the variance with self-efficacy adding a further 4.0% and the two variables sharing 6.1 % of the total variance. A similar pattern is apparent at the dimensional level with perceptions of success contributing 9.3-21.9% of the variance and self-efficacy continuing to be a significant but modest predic- tor of each dimension when controlling for perceptions of success (R2 = .014 to .075). The multiple regression results for perceived competence mirror the multivariate results with self-efficacy predicting more variance in this dimen- sion than in any other.

Discussion

The present study was designed to examine the relative strengths of the relationships between self-evaluative and self-efficacy mechanisms and in- trinsic motivation, at both the dimensional and composite levels of the con- struct, for aerobic exercise. As previously indicated, exercise and physical

Page 12: Self-Efficacy, Perceptions of Success, and Intrinsic Motivation for Exercise

150 MCAULEY, WRAITH, AND DUNCAN

Table 6

Hierarchical Regression Analyses Showing Unique and Cojoint Effects of Serf-Efficacy and Perceptions of Success on Overall Intrinsic Motivation and Dimensions

Predictor Unique Ffor Shared variable R’ change variance Dependent variable

Intrinsic Motivation Perceived success ,287 1 19.28

Self-efficacy .040 16.33

Total R2 = ,388, F(2,254) = 8 0 . 6 6 , ~ < .0001

Interest-Enjoyment Perceived success .170 55.76

Self-efficacy .O 17 5.69

Total R2 = ,221, F(2,254) = 3 6 . 0 7 , ~ < .0001

Perceived Competence Perceived success .I80 67.45

Self-efficacy ,015 27.94

Total R2 = .318, F(2,254) = 5 9 . 4 5 , ~ < .0001

Effort-Importance Perceived success .2 19 79.45

Self-efficacy .030 10.83

Total R2 = .297, F(2,254) = 5 3 . 6 8 , ~ < .0001

Pressure-Tension Perceived success .093 27.37

Self-efficacy ,014 4.24

Total R2 = .129, F(2,254) = 1 8 . 8 0 , ~ < ,0001

.ooo 1

.O 1

.ooo 1

.05

.ooo 1

.ooo I

. 000 I

.o 1

.001

.05

.06 1

.034

.063

.063

,022

activity are not readily adopted and maintained as a direct result of the intrinsic interest of the individual. Rather, exercise is often initiated for external reasons (e.g., health, appearance, etc.) and then, as physiological adaptation sets in and the individual perceives him/ herself to be self- determining in their activity, intrinsic motivation begins to play a role.

Bandura ( 1986) has proposed that high levels of efficacy denote competence and thereby bolster interest in the task. Conversely, deficits in perceived personal agency serve to attenuate intrinsic interest. Self-evaluative mecha- nisms such as goal attainments or perceptions of success are also hypothesized to be related to intrinsic motivation. The present findings offer some moderate support for these postulations.

By classifying subjects as having high or low exercise self-efficacy (based on

Page 13: Self-Efficacy, Perceptions of Success, and Intrinsic Motivation for Exercise

SELF-EFFICACY AND INTRINSIC MOTIVATION 151

extreme thirds) we were able to determine whether intrinsic motivation was affected by perceptions of personal capabilities. Highly efficacious individ- uals proved to be significantly more intrinsically motivated than their less efficacious counterparts. One might argue that these differences merely represent variations in experience or ability, as denoted by beginning, inter- mediate, and advanced status. Preliminary analyses suggested this not to be the case, with the three ability groups not differing significantly along the dimensions of intrinsic motivation. Thus, any differences along these dimen- sions appears to be partially due to the strength of the subjects’ exercise efficacy. Discriminant function coefficients suggested that, in spite of uni- variate differences at all dimensional levels of intrinsic motivation, it was perceived competence that maximally contributed to the multivariate effect. Bandura ( 1986) has previously indicated that self-efficacy plays an important role in mediating intrinsic motivation through perceived competence, a pat- tern suggested by the present results.

That perceptions of self-efficacy are moderately related to intrinsic moti- vation has been established. But what of the role played by self-evaluative mechanisms, in this case perceptions of success with respect to improved conditioning? How does this variable impact on the previously described relationship between efficacy and intrinsic motivation? Hierarchical multiple regression analyses examined the unique and cojoint effects (Pedhazur, 1982) of these variables on intrinsic motivation. These analyses revealed that both the self-efficacy and self-evaluative mechanisms were significant predic- tors of overall intrinsic motivation as well as at the dimensional level of the construct. However, the self-evaluative mechanism (perceived success) ac- counted for considerably more variance (Rz = .09 to .28) than self-efficacy (Rz = .014 to .09) with both variables sharing, on average, approximately 4.5% of the variance. That both mechanisms were in effect in the generation of intrinsic motivation is in keeping with the social cognitive perspective espoused by Bandura (1986). Although the self-evaluative mechanism was most strongly related to the intrinsic motivation variables, the self-efficacy mechanism appears to be serving an augmenting function. This is not unlike the recent model proposed by Vallerand ( 1987), in which intuitive appraisal (perceived success) plays a fundamental role in the generation of affect with other cognitions such as causal ascriptions augmenting these emotions. Clearly, further examination of the relationships between self-efficacy, per- ceived success, and other self-evaluative mechanisms and intrinsic motiva- tion is warranted. Furthermore, the present study assessed perceptions of success only with respect to improved conditioning. It is entirely probable that individuals may judge their success in an exercise program relative to other parameters such as weight loss or maintenance, mastery of technique, improved gracefulness, or in social comparison terms. Future attempts to

Page 14: Self-Efficacy, Perceptions of Success, and Intrinsic Motivation for Exercise

152 MCAULEY, WRAITH, AND DUNCAN

gauge the influence of perceived success would perhaps be wise to assess multiple aspects of such perceptions.

As the earlier multivariate analyses indicated, it was perceptions of com- petence that maximally contributed to the difference between high and low efficacy groups. That is, highly efficacious individuals appear to be more intrinsically motivated and this appears to be mediated by perceived com- petence. Indeed, the regression analysis confirmed this association with effi- cacy cognitions being more strongly related to the perceived competence dimension than any other dimension of intrinsic motivation.

Whether highly efficacious individuals remain intrinsically interested is, of course, dependent on the nature of the task or activity. It is highly question- able whether advanced aerobic exercise participants would continue to exhibit strong senses of intrinsic interest if the intensity, duration, complexity, and difficulty of the activity remained the same or was reduced. Indeed, having reached a certain level of performance and physiological functioning, challenge would appear to be a prerequisite for proficient individuals to continue in the pursuit of such activity.

Although the present results are interesting and suggest some important links among self-evaluative mechanisms, self-efficacy, and intrinsic motiva- tion for aerobic exercise, some notes of caution are appropriate. Unfortu- nately, the present design, being cross-sectional, did not allow for either analysis or interpretation that might speak to the causal associations among the variables of interest. Rather, the purpose of the investigation was to determine whether the hypothesized relationships did indeed exist in the exercise setting. Proposing that self-efficacy and perceptions of success influ- ence intrinsic interest in a task, however, makes some logical sense. After all, we can be intrinsically interested and engage in an activity without necessarily developing increased efficacy or perceptions of success. However, mastering a task that has held little interest previously can lead to the facilitation of intrinsic interest through increases in beliefs in personal capabilities and positive self-evaluation. This latter position is advocated by Bandura’s (1986) social cognitive theory. Now that a relationship between these variables in the exercise domain has been established it remains for future investigations to examine the lagged and longitudinal effects among variables over time. Cer- tainly, such a design would enable one to explore Bandura’s contention that the efficacy-intrinsic motivation link is subject to a temporal lag in its generation.

That perceived success and self-efficacy accounted for approximately 38% of the variance in intrinsic motivation warrants some attention. As previously noted, participants in exercise programs seldom commence the program out of heartfelt interest or the enjoyment of intense physical demands being placed on the system. Rather, the social cognitive approach (Bandura, 1986), which

Page 15: Self-Efficacy, Perceptions of Success, and Intrinsic Motivation for Exercise

SELF-EFFICACY AND INTRINSIC MOTIVATION 153

projects intrinsic interest to be gradually generated as the self-efficacy and self-evaluative mechanisms come into effect, appears to be a more salient interpretation of intrinsic interest in exercise behavior. The present data suggest that the above-mentioned variables are related to each other in the exercise setting. Such findings may have implications for issues concerning adherence to exercise and other health regimens. If mastery experiences at relatively demanding levels of task activity can be provided and, more impor- tantly, identified to the individual, such experiences coupled with attendant increases in self-efficacy appear likely to influence positively those aspects of intrinsic motivation that result in participation for less external and more internal reasons.

The present study suggests that self-evaluative and self-efficacy mecha- nisms may be implicated in the development of intrinsic motivation for aerobic exercise activity. If intrinsic motivation is indeed important for con- tinued participation in activity following initial adoption, then it would behoove researchers in this area to explore further the temporal relationships among thesevariables and how they might be manipulated or facilitated in the exercise setting to further contribute to our understanding of this very com- plex behavior.

References

Bandura, A. (1977). Self-efficacy: Toward a unifying theory of behavioral change. Psychological Review, 84, 19 1-2 15.

Bandura, A. (1986). Social foundations of thought and action: A social cognitive theory. New Jersey: Prentice-Hall.

Bandura, A., & Cervone, D. (1983). Self-evaluative and self-efficacy mecha- nisms governing the motivational effects of goal systems. Journal of Per- sonality and Social Psychology, 45, 1017-1028.

Bentler, P. M., & Bonett, D. G. (1980). Significance tests and goodness-of-fit in covariance structures. Psychological Bulletin, 88, 588406.

Bonnano, J. A., & Lies, J. E. (1974). Effects of physical training on coronary risk factors. American Journal of Cardiology, 33, 760-764.

Cooper, K . G. (1968). Aerobics. New York: Bantam Books. Cronbach, L. J. (1951). Coefficient alpha and the internal structure of tests.

Deci, E. L., & Ryan, R. M. (1985). Intrinsic motivation and self-determi-

Dishman, R. K . (1982). Compliance/ adherence in health-related exercise.

Dishman, R. K . (1986). Exercise compliance: A new view for public health.

Psychometrika, 16, 296-334.

nation in human behavior. New York: Plenum.

Health Psychology, 1,237-267.

The Physician and Sports Medicine, 14, 127-145.

Page 16: Self-Efficacy, Perceptions of Success, and Intrinsic Motivation for Exercise

154 MCAULEY, WRAITH, AND DUNCAN

Dishman, R. K. , Sallis, J . F.. & Orenstein, D. R. (1985). The determinants of physical activity and exercise. Public Health Reports. 100, 158-171.

Dzewaltowski, D. A. (1989). Toward a model of exercise motivation. Journal of Sport and Exercise Psychology, 11, 25 1-269.

General Mills Inc. (1979). The General Mills American family report 1978- 1979: Family health in an era of stress. New York: Yankelovich, Skelly, and White.

Joreskog, K . G., & Sorbom, D. (1984). LISREL: Analysis of linearstructural relationships by the methodof maximum likelihood. (Version VI) . Moores- ville, IN: Scientific Software.

Marsh, H. ( 1985). Application of confirmatory factor analysis to higher-order factor structures and factorial invariance. Multivariate Behavioral Re- search, 20,427-449.

McAuley, E . (1985). Success and causality in sport: The influence of percep- tion. Journal of Sport Psychology, 7, 13-22.

McAuley, E., Duncan, T., & Tammen, V. (1989). Psychometric properties of the Intrinsic Motivation Inventory in a competitive sport setting: A con- firmatory fact or analysis. Research Quarterl-y for Exercise and Sport, 60,

McAuley, E., & Jacobson, L. (in press). Self-efficacy and exercise participa- tion in sedentary adult females. American Journalof Health of Promotion.

McAuley, E., & Tammen, V. (1989). The effects of subjective and objective competitive outcomes on intrinsic motivation. Journal of Sport and Exer- cise Psychology, 11, 84-93.

Morgan, W. P. (1977). Involvement in vigorous physical activity with special reference to adherence. In G. I. Gedvilas & M. E . Kneer (Eds.), National College of Physical Education Association Proceedings (pp. 233-246). Chicago, IL: University of Illinois, Chicago.

National Center for Health Statistics (1980). Health United States, 1977- 1978. Washington, DC: U S . Government Printing Office.

National Institute of Mental Health. (1984). Coping with mental stress: The potential and limits of exercise intervention. Rockville, MD: National Institute of Health.

Oldridge, N. B. ( 1982). Compliance and exercise in primary and secondary prevention of coronary heart disease: A review. Preventive Medicine, 11,

Paffenberger, R. S. ( 1986). Physical activity all-cause mortality, and longevity

Pedhazur, E. J. (1982). Multiple regression in behavioral research (second

Perrier Great Waters of France, Inc. (1979). The Perrier study: Fitness in

49-58.

56-70.

of college alumni. New England Journal of Medicine. 314,605-613.

edition). New York: Holt, Rinehart, Winston.

America. New York: Perrier.

Page 17: Self-Efficacy, Perceptions of Success, and Intrinsic Motivation for Exercise

SELF-EFFICACY AND INTRINSIC MOTIVATION 155

Plant, R., & Ryan, R. M. (1985). Self-consciousness, self-awareness, ego- involvement, and intrinsic motivation: An investigation of internally con- trolling styles. Journal of Personality, 53,435449.

Ryan, R. M. (1982). Control and information in the intrapersonal sphere: An extension of cognitive evaluation theory. Journal of Personality andsocial Psychology, 45,736-750.

Ryan, R. M., Mims, V., & Koestner, R. (1983). Relation of reward contin- gency and interpersonal context to intrinsic motivation: A review and test using cognitive evaluation theory. Journal of Personality and Social Psy-

Sachs, M. L. (1984). Psychological well-being and vigorous activity. In J . M. Silva and R. S. Weinberg (Eds.), Psychological Foundations of Sport (pp. 435-444). Champaign, IL: Human Kinetics.

Siscovick, D. S. , LaPorte, R. E., & Newman, J. M. (1985). The disease- specific benefits and risks of physical activity and exercise. Public Health Reports, 100, 189- 195.

Spink, K . S., & Roberts, G . C. (1980). Ambiguity of outcome and causal attributions. Journal of Sport Psychology, 2 , 237-244.

U.S. Department of Health and Human Services. (1979). Surgeon General’s Report: Promoting health preventing disease, 1990 objectives for the nation. Washington, DC: Public Health Services, National Institutes of Health.

U.S. Department of Health and Human Services. (1980). Promoting health andpreventing disease. Washington, DC: Public Health Services, National Institutes of Health.

Vallerand, R. J. (1987). Antecedents of self-related affects in sport: Prelimi- nary evidence on the intuitive-reflective appraisal model. Journal of Sport

Vallerand, R. J., & Reid, G . (1984). On the causal effects of perceived competence on intrinsic motivation: A test of cognitive evaluation theory. Journal of Sport Psychology, 6,94-102.

Wankel, L. M. (1985). Personal and situational factors affecting exercise involvement: The importance of enjoyment. Research Quarterly for Exer- cise and Sport, 56, 275-282.

chology. 45,736-750.

Psychology, 9, 161-182.