• unlimited access with print and download
    $ 37 00
  • read full document, no print or download, expires after 72 hours
    $ 4 99
More info
Unlimited access including download and printing, plus availability for reading and annotating in your in your Udini library.
  • Access to this article in your Udini library for 72 hours from purchase.
  • The article will not be available for download or print.
  • Upgrade to the full version of this document at a reduced price.
  • Your trial access payment is credited when purchasing the full version.
Buy
Continue searching

Investigating the role of health value and the theory of planned behavior in predicting outcome expectancy and interest for physical activity in a worksite sample

Dissertation
Author: Bradley D. Benedict
Abstract:
The purpose of this study was to improve upon the theoretical and translational understanding of worksite wellness programming. The social cognitive constructs found within the Theory of Planned Behavior (i.e., perceived behavioral control, subjective norm, attitude) and ratings on the Health Value Scale were investigated in terms of their usefulness for predicting ratings on the Outcome Expectancy Scale and interest for physical activity. The sample was composed of staff and faculty members ( N = 98) of a small, private liberal arts university. Analyses to test the hypotheses were conducted by means of Pearson correlations, hierarchical multiple regression, and discriminant function analyses. The results did not reveal an improved discrimination between the self-efficacy and outcome value components of outcome expectancy. Attitude was significantly, positively correlated with health value and subjective norm was significantly, positively correlated with perceived behavioral control. Outcome expectancy was not significantly predicted by any of the research variables. However, a significant discriminant function was observed for outcome expectancy predicting gender. Variance in interest in physical activity was significantly explained by the age demographic in all steps of the regression model. Subjective norm significantly contributed to interest in physical activity when health value was entered into the regression model in the final step.

iv TABLE OF CONTENTS ACKNOWLEDGMENTS..................................................................................................ii ABSTRACT.......................................................................................................................iii LIST OF TABLES.............................................................................................................vi CHAPTER I.INTRODUCTION...................................................................................................1 Outcome Expectancy...............................................................................................3 Theory of Planned Behavior....................................................................................7 Examining Outcome Expectancy Using the Theory of Planned Behavior..............9 Examining the Role of Health Value as a Determinant of Outcome Expectancy.............................................................................................................11 The Theory of Planned Behavior as a Determinant for Physical Activity............12 Examining the Role of the Theory of Planned Behavior as a Determinant of Interest...............................................................................................................15 Examining the Role of Health Value as a Determinant of Interest........................15 Examining Gender Differences Among the Research Variables...........................16 Statement of Purpose.............................................................................................16 Proposed Analyses.................................................................................................16 II.LITERATURE REVIEW......................................................................................19 Outcome Expectancy.............................................................................................24 Theory of Planned Behavior..................................................................................32 The Theory of Planned Behavior as a Determinant for Physical Activity............34 The Role of Health Value within Previous Studies on Health Behaviors.............36

v The Role of Interest within Previous Studies on Physical Activity.......................38 III.METHOD..............................................................................................................41 Participants.............................................................................................................41 Measures................................................................................................................41 Procedures..............................................................................................................46 IV.RESULTS..............................................................................................................48 Sample Characteristics...........................................................................................48 Preliminary Analyses.............................................................................................48 Relationships Among the Research Variables.......................................................55 V.DISCUSSION........................................................................................................64 Review of Overall Findings...................................................................................64 Strengths of the Present Study...............................................................................71 Limitations to the Study and Directions for Future Research................................72 REFERENCES..................................................................................................................75 APPENDIX A:Study Questionnaire.................................................................................83 APPENDIX B:Consent Form...........................................................................................91 APPENDIX C:Drawing Form..........................................................................................92 APPENDIX D:Revocation Form......................................................................................93

vi LIST OF TABLES TABLE 1 Demographic Profile and Characteristics of Sample....................................49 TABLE2 Group Differences Between Males and Females Among Research Variables.......................................................................................................51 TABLE 3 Pearson Correlations Among Study Variables.............................................52 TABLE 4 Means and Standard Deviations Among Attitude Scale Items.....................52 TABLE 5 Intercorrelations Among Attitude Scale Items.............................................53 TABLE 6 Means and Standard Deviations Among OEE Scale Items..........................53 TABLE 7 Intercorrelations Among OEE Scale Items...................................................54 TABLE 8 Means and Standard Deviations Among Interest Scale Items......................55 TABLE 9 Intercorrelations Among Interest Scale Items...............................................57 TABLE 10 Hierarchical Multiple Regression Analysis Showing Amount of Variance in Outcome Expectancy Accounted for by Age,Theory of Planned Behavior Components,and Health Value.......................................58 TABLE 11 Hierarchical Multiple Regression Analysis Showing Amount of Variance in Interest in Physical Activity Accounted for by Age,Theory of Planned Behavior Components,and Health Value..................................60 TABLE 12 Standardized Function Coefficients and Correlation Coefficients of Study Variables Predicting Gender..........................................................62

1 CHAPTER I Introduction The current study attempts to contribute to the understanding and promotion of worksite wellness programming,defined as efforts within the workplace that enhance awareness,change behavior,and create environments that support good health practices (O’Donnell,1989).Selected social-cognitive variables that have been associated with previous physical activity research are examined with regard to their conceptualization, associations,and ability to predict interest for specific behaviors.The current project focuses on examining the components of outcome expectancy,the Theory of Planned Behavior,the ratings of health value,and demographic factors for predicting interest in specific physical activities introduced to participants as part of a potential worksite wellness programming effort.Therefore,the current work aims to accomplish research goals of providing basic theoretical contributions to the study of behavioral health change as well as results that are translational to improved design and implementation of actual worksite programming. Regular physical activity has been defined as “activity lasting at least 10 minutes that caused light sweating or a slight to moderate increase in breathing or heart rate” (U.S.Department of Health and Human Services [USDHHS],2003,p.2).It has been widely established that regular physical activity serves an influential role in the prevention and management of numerous physical diseases including coronary heart disease,cancer,and osteoporosis (Bouchard,Shephard,Stephens,Suttun,&McPherson, 1990;Harris,1980;USDHHS,1996).Physical activity has been widely reported to play a causal role in the management of body weight (USDHHS,2003) and to positively

2 correlate with the reduced incidence of work-related injuries (Goetzel,Guindon,Turshen, &Ozminkowski,2001;Warner,Wickizer,Wolfe,Schidroth,&Samuelson,1988).While physical inactivity has been associated with increased rates of depression (Strawbridge, Deleger,Roberts,&Kaplan,2003),exercise interventions can improve many aspects of self-esteem(Blair,Jacobs,&Powell,1985),emotional well-being,and quality of life (Arent,Landers,&Etnier,2000;McAuley &Rudolph,1995).Although the vast majority of American adults are aware of the health benefits of exercise (Morrow,Krzewinski- Malone,Jackson,Bungum,&FitzGerald,2004),it has been estimated that less than 15% of adults in the U.S.meet the Center for Disease Control guidelines for moderate physical activity (USDHHS,2003). The escalating health care costs related to employee disease,illness,and injury has led many public and private employers to introduce workplace initiatives to encourage employee physical activity as a way to contain those costs.Worksite wellness programming has been linked to reduced absenteeism,reduced health care costs,and increased worker productivity (Dunnagan,Haynes,&Noland,1999;Warner,et al., 1988).However,employee participation rates have been notoriously low (Dunnagan et al.,1999).In light of these findings,further empirical exploration regarding factors related to employee interest for physical activity,as well as alternative methods for predicting participation in worksite wellness programming,is warranted.Two commonly investigated theoretical models that have been found to play a useful role in understanding a wide range of behaviors in general,and physical activity in particular, are social cognitive theory (Bandura,1986,1997) and the theory of planned behavior

3 (Ajzen,1985,1991).Outcome expectancy (Bandura,1997) represents a portion of social cognitive theory that focuses on cognitions of self-efficacy and outcome value. Outcome Expectancy Social cognitive theory considers that environmental events,personal factors,and behavior are all interacting and reciprocal determinants of each other (Bandura,1986). With regard to behavior,this interplay between psychosocial variables within the context of a given situation influences how an individual may self-determine his or her participation.At the heart of this self-determination lies the content of one’s cognitions, thus,these beliefs,perceptions,or expectations serve a central role in Bandura’s model. Outcome expectancy is described as the belief or perception that a particular consequence will be produced by a certain behavior.Thus,outcome expectancy is considered to contain two fundamental elements (i.e.,one’s perceived ability to engage in the behavior and one’s belief about the outcome of the behavior). Self-efficacy.Self-efficacy,which is the belief or perception that one can perform a specific behavior,has been heavily applied to physical activity research and has been demonstrated to play an important role in physical activity,exercise,and other health- related behaviors (see Bandura,1986;McAuley 1992;O’Leary,1985;Sallis,Pinski, Patterson,&Nader,1988,for reviews).Bandura (1997) maintains that this perception of one’s confidence for accomplishing a given behavior largely determines the likelihood that the behavior will be initiated. Outcome value.While self-efficacy is considered to be a necessary feature of outcome expectancy,outcome value is also theorized to be a key factor and can be conceptually distinguished fromself-efficacy.According to Bandura (1997),in addition

4 to self-efficacy,it is also when an individual perceives that an outcome for a given behavior is positive or desirable,that he or she engages in that behavior.In contrast,the perception that a behavioral outcome is negative,or undesirable,supposedly reduces behavioral action.Likening positive outcome expectancy to values,outcome value moderates the effect of outcome expectancy for behaviors in that a valued positive outcome will lead to an engagement in behavior more than an outcome that is not valued. Relationships between outcome expectancy and self-efficacy constructs.A growing body of research has suggested some utility for predicting physical activity through the assessment of outcome expectancy (see Williams,Anderson,&Winett,2005 for a review).However,the extent to which this construct has been examined and deemed useful is far less than is the case for self-efficacy and the relationship between themhas been described as a source of perplexity and debate (Kirsch,1995;Maddux,1999; Resnick,2001a;Williams et al.,2005).Bandura’s conceptual depiction of outcome expectancy and self-efficacy indicates their necessary association,since self-efficacy is purported to be causally related to outcome expectancy.In this sense,an individual’s expected outcome regarding a particular physical activity behavior would be contingent upon a belief in his or her ability to performthe activity.Therefore,the relationship between self-efficacy and outcome expectancy would consistently reveal strong associations. Particularly troubling,though,is the tendency among the limited number of available physical activity research studies,examining both constructs,to report mixed results for empirically establishing a strong relationship between the two variables,or for providing a satisfactory empirical account of their differentiation.For example,some

5 studies have found that self-efficacy is related to outcome expectancy and that outcome expectancy does not provide improved prediction of physical activity beyond the influence of self-efficacy (Dzewaltowski,1989;Dzewaltowski,Noble,&Shaw,1990; Rovniak,Anderson,Winett,&Stephens,2002).Other studies,in contrast,report deviations between the factors and have suggested that outcome expectancy does serve to allow at least some improved prediction of physical activity beyond self-efficacy (Desharnais,Bouillon,&Godin,1986;Rodgers &Brawley,1996;Rodgers &Gauvin, 1998).Interestingly,the strongest of reported correlations between self-efficacy and outcome expectancy,where outcome expectancy shares the greatest variance with physical activity,has been reported in studies examining physical activity of older adults (Conn,1997;Resnick,2001a,2001b,2001c;Resnick,Orwig,Magaziner,&Wynne, 2002;Resnick,Palmer,Jenkins,&Spellbring,2000;Resnick,Zimmerman,Orwig, Furstenberg,&Magaziner,2000). Relationships between outcome expectancy and outcome value constructs.In their review of the utilization of outcome expectancy within physical activity research, Williams et al.(2005),suggest that older adults may value the positive outcomes of physical activity to a greater degree than younger people,leading themto engage in physical activity based more on outcome value than self-efficacy.Or,perhaps,outcome value for physical activity among younger adults may be present,but in ways that are not being measured.Regardless of which explanation may hold more merit,this example of apparent age differences,as well as the mixed reports of shared relationships and moderating effects,serve as illustrations that empirical attempts to gain a clearer

6 understanding of the factors complicating the outcome expectancy literature is preliminary,at best. In reviewing the available literature where the influences of outcome expectancy are studied in relation to physical activity,a particular approach is noticeably lacking.It is surprising that examples were not found where researchers have attempted to verify that the defining characteristic within outcome expectancy is that it actually does reflect positive or negative values for the outcomes that physical activity behaviors are expected to produce.It is thus useful to attempt to further establish the relationship between outcome expectancy and outcome value. Maddux (1999) and Williams et al.(2005) have pointed out that a major obstacle in addressing this issue in previous studies may lie with the fact that often,it is the nature of outcome expectancy measurements to contain both expectancy for engaging in a behavior (i.e.,self-efficacy) and expectancy that a behavioral outcome is valued (i.e., outcome value).In other words,one’s perceived likelihood that his or her engagement in a physical activity behavior will produce a particular result,and one’s perception that the result of that behavior is desirable,are incorporated within the same items.For example, the Outcome Expectations for Exercise Scale (OEE,Resnick,et al.,2000) includes items such as,“Exercise…Makes my muscles stronger,” and “Exercise…Helps to strengthen my bones” (Resnick et al.,2000,p.354).With this approach,it is unclear if an individual highly endorses the items because of his or her agreement that the likelihood of exercise accomplishes said results,or if he or she is indicating that they place a positive value on such results.On one hand,the outcome may be considered as likely and also valued.On the other hand,the same outcome could be perceived as likely,but not particularly

7 valued.In an attempt to help extricate the underlying contributing elements within outcome expectancy,and to provide a more clear differentiation with self-efficacy,the current work evaluates the outcome expectancy construct in relation to the well- established components found within another social-cognitive theoretical model of behavior change that utilizes constructs congruent to self-efficacy and outcome value, and perhaps in a more precise manner.Specifically,the role of self-efficacy and outcome value within outcome expectancy is evaluated within the context of the Theory of Planned Behavior (TPB;Ajzen,1985,1991,2002b) Theory of Planned Behavior The Theory of Planned Behavior (Ajzen,1985,1991,2002b) has been demonstrated to be useful in predicting a wide range of behaviors and has been specifically verified as useful for predicting behaviors related to exercise (Armitage & Conner,2001;Godin &Kok,1996;Hagger,Chatzisarantis,&Biddle,2002).The TPB identifies the conceptually independent constructs of attitude,perceived behavioral control,and subjective norm as principal determinants for behavior. Attitude.With regard to judgments about whether one values the outcome of specific behaviors,the attitude component reflects one’s positive or negative perceptions about performing a behavior and about the expected result of that behavior (Ajzen,1991). Therefore,the attitude construct within the TPB may be considered to be congruent with the theoretical depiction of the outcome value component within outcome expectancy. Items measuring attitude within the TPB require the respondent to indicate their strength of agreement between degrees of adjective extremes which allow the confidence of straightforward items tapping a perception of value regarding physical activity.

8 Examples of such polarized adjectives are “pleasant and unpleasant,” “good and bad,” and even “worthless and valuable” (Ajzen,2002a,p.5). Perceived behavioral control.Similar to the self-efficacy construct that is believed to be a causal ingredient within outcome expectancy,the perceived behavioral control component within the TPB is intended to assess one’s perceived ease or difficulty for performing a proposed behavior.For example,in response to the opening,“For me to regularly exercise would be…,” an individual indicates along a Likert-type scale their perception of the behavior being either “Impossible” or “Possible” (Ajzen,2002a,pp.6- 7). Subjective norm.Subjective normis designed to assess the perceived social pressure for an individual to performa behavior.This factor may be considered to reflect aspects of both self-efficacy and outcome value,since respondents consider the attitude of other “important” people regarding the potential behavior,and they may well reflect on whether that social feedback is desirable or not (Ajzen,2002a,pp.5-6). It is noteworthy that perceived behavioral control is the only TPB construct that is widely considered to directly influence the amount of variance that the TPB shares as a determinant of behavior (Armitage &Conner,2001).To explain,assuming that an individual has a high degree of positive attitude toward physical activity and they highly endorse perceived social pressure fromothers concerning physical activity,these contributions may still stop short of translating into actual behavioral performance.The reason for this is due to the possibility that a proposed physical activity may not completely reside under one’s volitional control (Ajzen,2002b).In other words,attitude and subjective normregarding a physical activity may be highly indicated,but perceived

9 behavioral control may not accompany these ratings.It is only in situations in which the proposed behavior is perceived to be under one’s control that the whole TPB model may allow predictive ability for behavioral action,and this contribution lends uniquely from the perceived behavioral control component. Examining Outcome Expectancy Using the Theory of Planned Behavior Examining associations between outcome expectancy and the Theory of Planned Behavior.As mentioned,outcome expectancy has been conceptualized to influence physical activity via its contributing elements of self-efficacy and outcome value. However,the physical activity literature appears to lack empirical attempts to demonstrate the degree to which the outcome expectancy construct actually reflects one’s perception that a behavior can be performed,as well as the extent to which that behavior is valued.It has been suggested that this fundamental establishment has been hindered largely due to the nature of existing measures for outcome expectancy to involve both self-efficacy and outcome value within the same items (Kirsch,1995:Maddux,1999; Williams et al.,2005) In an attempt to help distinguish the underlying contributing features within outcome expectancy,the current study proposes to examine its relationship with the TPB components.Specifically,the TPB components of perceived behavioral control and attitude may help to illuminate the role of self-efficacy and outcome value within outcome expectancy,since the perceived behavioral control construct is widely understood to reflect one’s self-efficacy for performing a behavior and the attitude construct is considered to be a straightforward assessment of one’s value for the behavior. If outcome expectancy heavily involves both self-efficacy and outcome value for

10 performing a behavior,we would expect a measurement of outcome expectancy (OEE, Resnick,et al.,2000) to be highly related to both the perceived behavioral control and attitude components of the TPB.If this is the case,we may also expect outcome expectancy to be highly correlated with subjective normwithin the TPB,since subjective normhas been considered to contain elements of both self-efficacy and value.However, if outcome expectancy is more associated with self-efficacy than outcome value,we would expect outcome expectancy to be more highly correlated with perceived behavioral control.In contrast,if outcome expectancy is more associated with outcome value,outcome expectancy should be more highly correlated with attitude than with perceived behavioral control within the TPB. Examining determinants of outcome expectancy.The current study also seeks to add to the findings within previous physical activity research involving outcome expectancy and the age of participants.In several studies,outcome expectancy was an important determinant of physical activity for older adults (Conn,1997;Resnick,2001a, 2001b,2001c;Resnick et al.,2002;Resnick et al.,2000;Resnick et al.,2000).It has been suggested that older adults may value the outcomes of physical activity to a greater extent than younger individuals (Williams et al.,2005).For the current study,actual physical activity is not assessed.However,it may be a worthwhile exploration to investigate whether participant’s age will serve as a determinant for scores on a measure of outcome expectancy (OEE,Resnick,et al.,2000).If older adults place greater value on the outcomes of physical activity,we would expect older individuals within the study to have higher ratings of outcome expectancy.

11 Beyond an examination of the correlational relationships between outcome expectancy and the TPB and the determinant ability of age for predicting outcome expectancy,the current study attempts to clarify the degree to which either self-efficacy or outcome value,as measured by the TPB,contributes to endorsements of outcome expectancy.Further,the extent to which the TPB contributes to the prediction of outcome expectancy above and beyond that which is accounted for by age will be examined. Therefore,the TPB components of perceived behavioral control,attitude,and subjective normwill be used as criterion variables to predict outcome expectancy.If Bandura’s (1997) conception that outcome expectancy is largely influenced by self-efficacy and outcome value,we would expect the TPB components of perceived behavioral control and attitude to both be predictors of outcome expectancy (OEE,Resnick,et al.,2000). Further,if self-efficacy is the primary causal ingredient for outcome expectancy,we would expect the TPB component of perceived behavioral control to contribute the greatest variance with outcome expectancy. Examining the Role of Health Value as a Determinant of Outcome Expectancy Another area of interest within the current study is to examine the extent to which the value that one places on their health may influence their ratings of outcome expectancy as measured by the OEE Scale (Resnick,et al.,2000) above and beyond that which may be accounted for by age and the TPB.This area of examination may provide some additional conceptual understanding about the underlying contributors to outcome expectancy.As mentioned,the TPB attitude items tap perceptions of outcome value by using straightforward pairs of adjectives extremes (i.e.,pleasant-unpleasant,harmful- beneficial,good-bad,worthless-valuable,enjoyable-unenjoyable).What is unclear,

12 though,is whether one’s endorsements of high or low attitude are representative of his or her value considerations for exercise behavior itself,or the value that they may place on the health outcome of exercise. Similarly,on one hand,the OEE Scale appears to inquire about potential health- related outcomes for exercise (i.e.,feel better physically,feel less tired,makes my muscles stronger,improves my endurance,strengthens my bones).On the other hand,all of these statements are presented in the context of one’s self-efficacy for exercising to attain these outcomes.The degree to which individuals value outcomes versus their self- efficacy for achieving the outcome through their own action is still in question.Thus,for the current study a measure of general health value is introduced.If general value for health is an important feature within of outcome expectancy for exercise,as opposed to self-efficacy for exercise or outcome value,then health value,as measured by the Health Value Scale (Lau,et al.,1986),will be demonstrated to have unique predictive variance with scores on the OEE above and beyond that accounted for by age and TPB. The Theory of Planned Behavior as a Determinant for Physical Activity Beyond an investigation of the psychometric associations between measurements of social-cognitive variables,the purpose of the current study is to examine the transportability of results to actual worksite programming efforts to increase participation in physical activities.It is noteworthy that the participants in the current study are asked to rate their interest for participating in physical activities.Actual physical activity behavior is not assessed. The body of research using the TPB for predicting physical activity/exercise behavior is vast.In many cases,the TPB is used to predict the intention to participate in

13 physical activity,rather than actual physical activity,itself.Accordingly,Ajzen (1985, 1991,2002b) posits that one’s intention to engage in a particular behavior serves as the most principal determinant for that behavior.Consequently,intention serves as a composite construct derived fromthe conceptually independent TPB components of attitude,perceived behavioral control,and subjective norm.Several studies have reported that the TPB explains significant variance for both intention to performphysical activities and actual physical activity behavior (Armitage &Conner,2001;Godin &Kok,1996; Hagger et al.,2002). Armitage and Conner (2001) provide results fromtheir meta-analysis reviewing the efficacy of the TPB froma database of 185 independent studies published through 1997.Overall,the TPB accounted for 27%of the variance for predicting behavior,and 39%for intention.Thus,the TPB was found to do a more effective job at predicting intention,rather than behavior,per se.To the degree that the constructs could be examined as independent contributors,perceived behavioral control provided the greatest amount of variance in both intention and actual behavior.In contrast,the subjective norm component was found to be a weak predictor of both intention and behavior. Likewise,Hagger and colleagues (2002) used meta-analytic techniques to examine 72 studies,dated as recently as 2001,that used the TPB components for predicting physical activity.When all TPB components were considered,45%of the variance for intention,and 27%for behavior was explained.Again,perceived behavioral control,as well as attitude,accounted for the greatest variance in both intention and behavior,with perceived behavioral control serving as a slightly more robust predictor

14 for both.Subjective normcontributed a small,but significant,influence on intention,but not behavior. Similarly,in their review of 18 studies using the TPB for examining exercise, Godin and Kok (1996) reported that attitude and perceived behavioral control served as the strongest and most consistent determinants of both intention and actual exercise behavior.Among their reviewed studies,intention and perceived behavioral control accounted for approximately 36%of the variance in exercise behavior.However, intention was reported as a significant determinant in all studies,and perceived behavioral control was shown to be a significant determinant in about half of the studies. Interestingly,a review of the available physical activity literature using the TPB does not produce examples in which researchers have examined the usefulness of considering participant’s interest for participation in physical activities.Rather,the measurement of intention to participate in physical activity,or actual physical activity behaviors,are consistently the variables of focus when investigating the determinant ability of the TPB constructs (i.e.,subjective norm,attitude,and perceived behavioral control).With regard to research involving physical activity programming for worksites, this is remarkable considering the expenses of finances,time,and effort inherent in the development and provision of worksite physical activity programs.Adding to this,since the employee participation rate for such programs have been historically low,the possible value of predicting interest may be a worthwhile investigation as a pragmatic first step. Specifically,if the TPB were to be demonstrated as a meaningful determinant of interest to participate in worksite-based physical activity programming,then developers of such programs may benefit fromexamining participant interest before devoting further

15 resources.However,it is not within the scope of the current study to establish the relationship between interest and intention,or the variance that interest shares with actual physical activity behaviors. Examining the Role of the Theory of Planned Behavior as a Determinant of Interest In order to determine if the TPB serves as a useful model for determining interest in physical activities,the TPB constructs of attitude,subjective norm,and perceived behavioral control are used to determine whether these variables are important in predicting interest to engage in specific health behaviors.If the TPB contributes to the prediction of interest in physical activity as it has been demonstrated to predict intention and actual physical activity,then we would expect all components of the TPB to provide significant variance with interest.Further,since perceived behavioral control has been widely reported in previous research as the most influential contributor to intention and physical activity,we may expect perceived behavioral control to also provide the greatest contribution as a determinant of interest.In addition,it may be a worthwhile exploratory inquiry to also assess the predictive variance that participant’s age shares with interest for physical activity.Therefore,age will be examined as a predictive variable within the regression model and the TPB will be investigated in regard to whether the TPB provides predictive variance in interest above and beyond participants’ age. Examining the Role of Health Value as a Determinant of Interest Similar to the method described for examining the role of health value in predicting outcome expectancy,the current study also investigates the extent to which the value individuals place on their health serves as a determinant of interest in physical

16 activity.Therefore,in addition to role of the TPB to predict interest in physical activity, health value is assumed to add predictive value to interest for physical activity. Examining Gender Differences Among the Research Variables In order to determine if the social cognitive variables used within the current study differ with regard to gender,a discriminant function analysis will be utilized.This analysis will demonstrate whether differences between male and female variable means allow a prediction of gender group membership.Therefore,the measures of outcome expectancy,health value,perceived behavioral control,attitude,subjective norm,and interest in physical activity will be entered into the discriminant equation to predict gender. Statement of Purpose The purpose of the current research is to assess social-cognitive variables associated with physical activity research in order to promote and enhance the future development of worksite wellness programming. Proposed Analyses Statistical analyses will be performed using the Statistical Package for the Social Sciences,Release 14.0 (SPSS,2003).Analyses will be conducted to determine whether there are any basic demographic differences among the components of the TPB and health value,and any significant differences will be entered as covariates. Hypothesis 1.It is predicted that outcome expectancy,as measured by the OEE Scale,will be significantly positively correlated with attitude,subjective norm,and perceived behavioral control and outcome expectancy will be demonstrated to have the highest significant positive correlation with perceived behavioral control.

17 To test this hypothesis,Pearson correlations will be conducted with outcome expectancy and the TPB (i.e.,subjective norm,attitude,and perceived behavioral control) entered as matrix variables to examine the relationship between the variables. Hypothesis 2.It is predicted that components of the TPB (i.e.,perceived behavioral control,attitude,subjective norm) will predict outcome expectancy above and beyond that which is accounted for by age.Within the TPB,perceived behavioral control will provide the greatest contribution to the variance with outcome expectancy. To test this hypothesis,a hierarchical multiple regression analysis will be conducted with the OEE scale entered as the dependent variable.Age will be entered as a predictor variable in step 1 of the model.In step 2 of the model,the TPB (i.e.,perceived behavioral control,attitude,subjective norm) will be added as predictive variables. Hypothesis 3.It is predicted that health value,as measured by the Health Value Scale,will provide unique predictive variance to scores on the OEE Scale above and beyond that accounted for by age and the TPB (i.e.,perceived behavioral control, attitude,subjective norm). To test this hypothesis,an additional step will be added to the hierarchical multiple regression analysis proposed to test hypothesis 2.In the third step,ratings from the Health Value Scale will be added to the model as a predictive variable. Hypothesis 4.It is predicted that the TPB (i.e.,perceived behavioral control, attitude,subjective norm) will predict variance in ratings of interest for physical activity above and beyond that accounted for by age.Within the TPB,perceived behavioral control will provide the greatest contribution to the variance with interest in physical activity.

18 To test this hypothesis,a hierarchical multiple regression analysis will be conducted with the total scale for interest in physical activity entered as the dependent variable.In step 1 of the model,age will be entered as predictive variable.In step 2 of the model,the TPB (i.e.,perceived behavioral control,attitude,subjective norm) will be added as a predictive variable. Hypothesis 5.It is predicted that health value,as measured by the Health Value Scale,will provide unique predictive variance to ratings of interest for physical activity above and beyond that accounted for by age and the TPB (i.e.,perceived behavioral control,attitude,subjective norm). To test this hypothesis,an additional step will be added to the hierarchical multiple regression analysis proposed to test hypothesis 4.In the third step,ratings from the Health Value Scale will be added to the model as a predictive variable. Hypothesis 6.It is predicted that components of the TPB (i.e.,perceived behavioral control,attitude,subjective norm),health value,as measured by the Health Value Scale,outcome expectance,as measured by the OEE Scale,and interest in physical activity will predict gender in the current sample. To test this hypothesis,a linear discriminant function analysis will be applied to the independent variables of outcome expectancy,the TPB components (i.e.,perceived behavioral control,attitude,subjective norm),health value,and interest in physical activity to determine whether any individual independent variable or combination of variables reliably predict gender.

Full document contains 104 pages
Abstract: The purpose of this study was to improve upon the theoretical and translational understanding of worksite wellness programming. The social cognitive constructs found within the Theory of Planned Behavior (i.e., perceived behavioral control, subjective norm, attitude) and ratings on the Health Value Scale were investigated in terms of their usefulness for predicting ratings on the Outcome Expectancy Scale and interest for physical activity. The sample was composed of staff and faculty members ( N = 98) of a small, private liberal arts university. Analyses to test the hypotheses were conducted by means of Pearson correlations, hierarchical multiple regression, and discriminant function analyses. The results did not reveal an improved discrimination between the self-efficacy and outcome value components of outcome expectancy. Attitude was significantly, positively correlated with health value and subjective norm was significantly, positively correlated with perceived behavioral control. Outcome expectancy was not significantly predicted by any of the research variables. However, a significant discriminant function was observed for outcome expectancy predicting gender. Variance in interest in physical activity was significantly explained by the age demographic in all steps of the regression model. Subjective norm significantly contributed to interest in physical activity when health value was entered into the regression model in the final step.