- Full ... t A A I l o ‘ ’0 n t u - . .. 1|... ‘ t .. vv. I... .u‘I- Itzv“l.lu..1 onlvn- - . I .It in. ‘11.. £3. a a . ‘ o ia‘zileI‘o‘ti AL v t I. I YAIO O A l . a u [31' 50.10 n.b;.f.l.l 57$ 2.." til}... EDISQ‘ v-0 3 . tullh“! » I : “fill" .. v.3“. -\ Alflnw'lvhi. Inv‘ln‘L Lk A ' . n i v V 1”“89} you | “HWW’IO . A. o ov - I... ..|| . t‘ :1 ~ v 5th”. I .p..v|l. . ‘ n I: l 4 I! n! I. fl .c t v. I ‘ \ .vt ‘ an t I v It. ‘ L. : fl 9 o yo avuv1 I t: . ¢ .‘I Q o! c .050. .o In; I: .\ . 7. \II o v 0.- ~ v I. T! .0 0 2 .7 .. u 1.. v 4 u Aunt -01.. . It . .1 . 13%|. La 0 '1 I n o I n 1 I .. v.1 .A v 2! ul 9 .1 \ .v w I .t. v. . . . t. .r .. I | ‘<‘ 7 c . 4hfihul|v~ a. 21“...» t 1...- “la n. 1"!!! ‘ . .. .......u+1txu.; b? L .zilu. . -. - :13?! s -i .Lluvt J” V‘ 0 . ' IIA . . - ' A n o I)‘ II. L N! O. I I l- ,l. i ? . 1 . lnl :0. gilt!!!) who," I! ulnflr.|.ihfiuu . |..# I? . X A I . I?!) . .1 infill-s :‘IIUJHVV i‘ir‘n‘flfiblfil“r\'ulniau‘i)flu| ‘lfli Ill. '1‘, Iv‘ufi . rule! , 1.1} 5““. a (fret. :1. -t' It} .. .ftpliot... .l --n{.l {III--11).! LI 3.1.2 isf'ulllrl an dd... W 8 ‘Rié’lr’ 1.113.151.1033}... il'r’nfluumlil -1 llszlv- .vtllllbxv( l. . 3 . luv}... . , ”hum-Pd» 1:... 1| .l‘fi: . ii 1 I10... 4 iiighunntl 1.1 11. {giltutwfl .tl l. .. (8:31.. ,1 . . . “ V‘ti L In t ll‘ . I .l‘llr|\ I Is (ts); Iii“ p . A iflfl‘: Ian“! cl. it (If. n\ I 1 a | Ir 'li if C " i I 1.. . . . 923Vifiu‘qflu‘1c’ll“, sv‘li.i.‘l|lt ‘ I; . I F . . :l‘l .. J . . \I} ,t.“t\|l\él 1110,44 03]: .111" u.\. '19.... I" o .4}. . 1 I , ... .. .Yl.. l :— .‘|ICI¢|I’ I .1113)! ’40.! ‘ v vPcAl‘iot.’l. Lira.)~14n\l‘ti1¢l4t (11] ‘10? . ll \l ,\‘ A no .. ".1. .I p v ) ..«. . ‘1: , .' ...v inf t v 4‘ P \ (L |. .91.. Evilgtdfui. .ul.v1..tl .... .) Ida. «It‘lpflvflcltvl‘:‘.4?4Xt1|.\llll{\.¢l|:14nhlvtp1o!ukl”‘x‘lvtuf‘lo t'tNHva.A|.Q.'IO . ... I .. :. . . . ... u... .:,. _ r . . .. .. .u . v . .n. 1...... ..‘,I.‘ 09!! I). «‘1. tit}; .Y.\4.,V4I3.1ov ...l (4. .o . Onavl... 3:14.} .41 v :( (la. V . 4.. . K. .1 I“. .. .. . .x . 3. [34:1‘ .LwIA 01.. . .I ‘ « ‘ .A .11 vli’l’trl‘v.‘ lz)1.|1« ‘SDIIO‘ llulccvu? I- t1 . ‘ . .V, . .1 1 I v. V 0.: .A.I\ .1» put. . ...v.... n 1‘: . .. .. r. . .4 . n a 3. ill L NIS.‘ 4.-.», t¥.r.\-nl..v.l. .‘l‘v n !. nil]. a. . THESlS ‘\ This is to certify that the thesis entitled - ....; .5‘. . -....-. 0- ,_.—... . -~ --—~-- THE INFLUENCE OF MODEL GENDER, MDDEL :-ir PERCEIVED ABILITY AND SUBJECT PERCEIVED_ _Hmafi ABILITY ON MUSCULAR ENDURANCE AND sELE— ”““ ’ EFFICACY ._ .-. . .- “nu“..- - presented by” JANIE SPREEMAN’ has been accepted towards fulfillment of the requirements for Master Of Arts degree in Physical Education Wiw Major professor Date 5/19/83 0-7639 MS U is an Affirmative Action/Equal Opportunity Institution PLACE IN RETURN BOX to remove this checkout from your record. TO AVOID FINES return on or before date due. DATE DUE DATE DUE DATE DUE ‘ m5 Emmi}? L—_______-——‘ l.____. MSU Is An Affirmative ActiorVEqual Opportunity Institinion cmmuna-q “(7’ -’/'j7/ r A ‘ . "‘3‘ f. 7 ( \— I 7"“ a W ' THE INFLUENCE OF MODEL GENDER, MODEL PERCEIVED ABILITY AND SUBJECT PERCEIVED ABILITY 0N MUSCULAR ENDURANCE AND SELF-EFFICACY By Janie Spreemann A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF ARTS Department of Health and Physical Education 1983 ABSTRACT THE INFLUENCE OF MODEL GENDER. MODEL PERCEIVED ABILITY AND SUBJECT PERCEIVED ABILITY ON MUSCULAR ENDURANCE AND SELF-EFFICACY By Janie Spreemann This study was designed to investigate the influence of model/ observer similarity cues (gender and perceived task ability) on model- ing and self-efficacy. This study employed a Model Gender by Model Athletic Ability by Subject Athletic Ability factorial design plus an external control group for each subject ability group. College-aged female athletes and nonathletes (N = 150) observed a videotape of either a female or male, athletic or nonathletic model perform a leg- endurance task, or were assigned to a no-model condition. Each subject was required to sit on a stool and extend her leg above a cord. _The length of time the leg remained extended on three trials was recorded. Pre- and post-experimental queStionnaires determined a) the effective- ness of model similaritYI’dissimilarity manipulations, and b) subjects' self-efficacy. Results indicated that athletes performed significantly longer and had stronger efficacy expectations than nonathletes. Subjects obserVing a female athletic model tended to perform better than subjects observing a male athletic model, although this difference was not significant. Athletes perceived themselves as significantly more similar to the athletic model, while nonathletes thought they were more similar to the nonathletic model. Both groups thought they were more similar to the female than male models. ACKNOWLEDGEMENTS I wish to extend my sincerest gratitude to my advisor, Dr. Deborah L. Feltz, for her unyielding energy, supportive encouragement, continual guidance, and thorough dedication to this thesis. Her door was always open to my numerous and frequent questions, regardless of how trivial, confusing or time-consuming they may have been. The efficiency, organi- zational abilities, countless hours of review and determination of Dr. Feltz, helped me to meet my deadlines and achieve my many goals. I wish to thank my former advisor and committee member, Dr. Daniel Gould, for his invaluable advise, creative suggestions and unending sup- port. His influence will always be remembered and greatly appreciated, not only in his efforts towards the completion of this thesis, but in his sincere concern with my well-being and growth as an individual and educator. The impression that Dan Gould has left on me will remain with me and inspire me throughout the rest of my life. I also wish to thank Dr. Luke Kelly for his invaluable feedback and advice as a member of my thesis committee. Dr. Kelly's insight and concern provided me with a well-rounded outlook regarding the purposes and implications of this study. Appreciation is extended to Dr. Maureen Weiss for her suggestions and "helpful hints"; my models, Denise Mugno, Gregg Parini, Tina Van Dommelen and Chris King; Bernie Holland and Dr. Crystal Branta for the ii use of their classes for my manipulation checks; and Rich Kimble for his technical assistance. Special thanks are in order for Denise Mugno, Pete Girolamo and my mother, for their continual support, encouragement, and understand- ing of the rigors of "graduate school life". Lastly, I am indebted to my mother for her countless hours of "long-distance" typing skills which were invaluable to the completion of this work. TABLE OF CONTENTS CHAPTER Page LIST OF TABLES ....................................... vi LIST OF FIGURES ...................................... viii I. INTRODUCTION ......................................... 1 Nature of the Problem ............................. 1 Statement of the Problem .......................... 9 Hypotheses ........................................ lO Delimitations ..................................... 10 Definitions ....................................... 10 Limitations ....................................... ll II. REVIEW OF THE LITERATURE ............................. l3 Observational Learning Function of Modeling ....... 16 Response Facilitation ............................. 20 Inhibition/Disinhibition of Responses ............. 21 Model/Observer Similarity ......................... 24 Perceived similarity and proficient performers. 34 III. METHODS AND PROCEDURES ............................... 37 Subjects and Design ............................... 38 Experimental Task ................................ 39 Model Similarity/Dissimilarity Manipulations ...... 4l Manipulation Check ................................ 43 Questionnaires .................................... 45 Pre-experimental questionnaire ................. 45 Post-experimental questionnaire ................ 45 Pilot Study ....................................... 46 Procedure ......................................... 47 Testing environment ............................ 47 Administration of test ......................... 48 Treatment of Data ................................. 52 iv CHAPTER Page IV. RESULTS ............................................... 54 Motor Performance Results .......................... 55 Hypothesis 1 .................................... 61 Hypothesis 2 .................................... 61 Hypothesis 3 .................................... 62 Questionnaire Results .............................. 62 Self-efficacy results ........................... 53 Model/observer similarity results ............... 73 Competition, influence and comparison results... 76 Background information results .................. 79 Summary of ANOVA Results ........................... 80 V. DISCUSSION AND CONCLUSIONS ............................ 83 Discussion ......................................... 84 Conclusions ........................................ 92 Suggestions for Future Research .................... 92 REFERENCE NOTES .............................................. 94 REFERENCES ................................................... 95 FOOTNOTES .................................................... lOO APPENDICES A. PRE-EXPERIMENTAL QUESTIONNAIRE ........................ 101 B. POST-EXPERIMENTAL QUESTIONNAIRE ....................... 103 C. MOTOR PERFORMANCE DATA ................................ 105 D. QUESTIONNAIRE DATA .................................... 107 E. RAN DATA ......... . ..................................... 122 TABLE 10. 11. 12. LIST OF TABLES . Means, Standard DeViations and Discriminant Function Coefficients for the Ten Model Characteristics by Model Type ............................................ . Means and Standard Deviations for Motor Performance of Experimental versus Control Subjects .................. . Analysis of Variance for Motor Performance of Experi- mental-Only Subjects .................................. . Analysis of Variance for Motor Performance of Experi- mental versus Control Subjects ........................ . Analysis of Variance for Questionnaire Item 3--Level of Efficacy for Experimental-Only Subjects ............ . Analysis of Variance for Questionnaire Item 3-- Strength of Efficacy for Experimental-Only Subjects... . Analysis of Variance for Questionnaire Item 4--Level of Efficacy for Experimental-Only Subjects ............ . Analysis of Variance for Questionnaire Item 4-- Strength of Efficacy for Experimental-Only Subjects... . Analysis of Variance for Questionnaire Item 3--Level of Efficacy for Experimental versus Control Subjects.. Analysis of Variance for Questionnaire Item 3-- Strength of Efficacy for Experimental versus Control Subjects .............................................. Analysis of Variance for Questionnaire Item 4--Level of Efficacy for Experimental versus Control Subjects.. Analysis of Variance for Questionnaire Item 4-- Strength of Efficacy for Experimental versus Control Subjects .............................................. vi Page 44 56 105 106 107 108 109 110 111 112 113 114 TABLE Page 13. Analysis of Variance for Questionnaire Item 5--Leve1 of Efficacy for Experimental-Only Subjects ............ 115 14. Analysis of Variance for Questionnaire Item 5-- Strength of Efficacy for Experimental-Only Subjects... 115 15. Analysis of Variance for Questionnaire Item 5--Leve1 of Efficacy for Experimental versus Control Subjects.. 117 16. Analysis of Variance for Questionnaire Item 5-- Strength of Efficacy for Experimental versus Control Subjects .............................................. 118 17. Analysis of Variance for Questionnaire Item Assessing Strength of Efficacy for 15 Time Lengths of Experi- mental-Only Subjects .................................. 119 18. Analysis of Variance for Questionnaire Item Assessing Strength of Efficacy for 15 Time Lengths of Experi- mental versus Control Subjects ........................ 120 19. Analysis of Variance for Questionnaire Item Assessing Model/Observer Perceived Similarity Rating for ExPerimental-Only Subjects ............................ 121 vii LIST OF FIGURES FIGURE Page 1. Experimental leg-lift apparatus ....................... 4o 2. Testing room .......................................... 49 3. Model Gender by Model Athletic Ability Interaction on performance ........................................... 57 4. Model Gender by Model Athletic Ability by Subject Athletic Ability Interaction on performance ........... 59 5. Model Gender by Subject Athletic Ability by Pre/Post Measures Interaction for efficacy strength on Item 5.. 69 viii CHAPTER I INTRODUCTION Nature of the Problem In many languages, the word for teach has the same meaning as the word for show (Reichard, 1938). This statement seems to have met with general acceptance in the physical education and athletic domain where the concept of modeling has become an inherent and focal part of the instructional system. Modeling can certainly facilitate the develop- ment of motor behavior by decreasing the inefficiency of trial and error learning that is often associated with the development of finely coordinated and complex sequences of motor skills. One of the most plausible and encompassing theories of modeling appears to be the stimulus-contiguity or social learning theory of modeling proposed by Bandura (1969). Modeling, which refers to behavior that results from an observer's exposure to another individual's per- formance or behavior, is conceptualized by Bandura as primarily an informational process whereby the observer is cognitively engaged as a central information processor. That is, people process, weigh, and integrate diverse sources of information regarding their capabilities, and they regulate their choice behavior and effort expenditures accord- ingly (Bandura, 1977a). In order to effectively integrate the.Various aspects of the modeling process into a common conceptual framework, Bandura has identified four basic components which are essential to modeling theory. Specifically, these four components are attention, retention, motoric capability and motivation. For an observer to sub- sequently display the modeled act, he or she must first attend closely to and retain what was demonstrated. In addition, the observer must possess the motoric capability to reproduce the modeled act and be motivated to do so. The modeling process functions in several ways. According to Bandura (1977b), modeling influences can serve as instructors, inhibit- ors, disinhibitors, facilitators, stimulus enhancers and emotion‘ arousers. In conjunction with this notion, Bandura (1971) has concep- tualized three primary modeling functions. The first of these functions, observational learning, is mainly concerned with acquiring novel response patterns as the result of observing a model. Much of the modeling interest in physical education deals with the research and application of observational learning concepts. Motor behavior research dealing with observational learning has concentrated heavily on strategy- oriented laboratory tasks, such as the "shoot-the-moon" game and the Bachman ladder climbing task (Landers & Landers, 1973; Martens, Burwitz & Zuckerman, 1976). The second function of modeling, response facilitation, is char- acterized by cues obtained as a result of observing a model perform non- novel, nonthreatening tasks. In this process, cues serve to enhance the performance of pre-existing responses (Bandura, 1971). For example, if a coach wanted to increase players' praising of their teammates, he might have the team captain model the desired behavior. This behavior already exists in the players' repertoire, but the team captain's behavior facilitates the response. The third function of modeling, the inhibitory/disinhibitory function, deals with performance decrements or increments as the result of observing a model's performance, as well as the subsequent conse- quences associated with that performance. The imitation of modeled behaviors are classified as inhibitory when the observer shows a decre- ment or general reduction of responsiveness as a result of seeing a model's behavior produce punishing or aversive consequences (Bandura, 1971). Examples of research investigating the inhibition effect are reflected by various studies centering on the topic of aggression (Bandura, 1973; Bandura & Walters, 1959). In contrast, disinhibitory effects occur when observers increase the performance of formerly in- hibited behaviors as a result of observing models engage in threatening, persevering, or prohibited activities without experiencing aversive consequences (Bandura, 1971). Research investigating disinhibition effects of modeling has focused primarily on phobic reactions and avoid- ance behavior exhibited toward snakes (Bandura & Adams, 1977; Bandura & Barab, 1973; Bandura, Blanchard & Ritter, 1969; Borkovec, 1973; Kazdin, 1973, 1974; Kornhaber & Schroeder, 1975; Meichenbaum, 1971). Another application of the disinhibition effect seems highly relevant to the field of physical education and athletics where fear, persistence, pain, and endurance are often evidenced in many physical activities. Although modeling tasks which require perseverance and endurance have not been traditionally categorized under the rubric of disinhibition modeling effects, the fact that these tasks are experi- enced by some discomfort, would seem to qualify them under this category. A paucity of this type of disinhibition research directly related to the field of physical education and athletics exists, however, with only a few published studies to date examining this phenomenon (Feltz, Landers & Raeder, 1979; Gould & Weiss, 1981; Lewis, 1974). Regardless of the type of modeling effects examined, two areas receiving a considerable amount of attention in the modeling research have been the attentional and motivational subprocesses of Bandura's (1969) theory. Of particular interest have been the influencing factors of both model and observer characteristics. Various manipula- tions of both model and observer characteristics have been employed in an attempt to define the optimal model/observer characteristics neces- sary for the enhancement of modeling. Research of this nature has focused upon such issues as competence, status, and similarity of the model. For instance, in the area of observational learning, it has been generally concluded that a model of higher competence, Skill, and social status will more positively affect the observer‘s attention to the model and motivation to perform than will a less competent, lower skilled, and lower social status model (Bandura, 1969; Bandura & Walters, 1963; Baron, 1970; Flanders, 1968; Landers & Landers, 1973; Rosenbaum & Tucker, 1962; Zimmerman & Blotner, 1979). In the area of inhibition/disinhibition research, the effects of various dimensions of model/observer similarity have become an integral focal point. According to Bandura (_1977a), model similarity increases the personal relevance of the modeled performance to the observer and, therefore, should increase the observer's motivation to perform. Model similarity exists along a number of dimensions such as gender, task ability, age, race, and I.Q. For example, the dimension of gender has been investigated by employing models of similar or dissimilar gender to the observer, in hopes of creating an enhanced condition of perceived similarity or dissimilarity between model and observer. The results of these investigations suggest that enhanced treatment effects become evident when similar-gender models are employed (Gould & Weiss, 1981; Kazdin, 1974; Perry & Perry, 1974; Rosekrans, 1967), although these effects have been confounded with other variables. Also, despite numerous conflicting reports related to model/observer similarity by gender effects (Flanders, 1968), most researchers seem to advocate the use of same-gender models when attempting to enhance the modeling process. Hence, much of the modeling research, regardless of academic discipline, has characteristically employed same-gender models. Perceived similarity of the model (in terms of task ability or perceived ability in competence) has been another variable of interest in the inhibition/disinhibition function of modeling. However, this variable has been historically confounded by the inclusion of same-gender models in the majority of research designs, making it difficult to demon- strate the causality between perceived similarity in task ability and enhanced modeling effects. For instance, disinhibition studies investi- gating approach behavior by snake phobic subjects have examined per- ceived similarity of the model in terms of fear and its effect on subse- quent approach behavior. The models used in these studies have always been of the same gender as the observer. Generally, it has been concluded that subjects perceiving themselves as manifesting similar fear and approach behavior as the snake phobic model, displayed an increase in approach behavior and a decrease in fear as the result of observing this similar model cope with the situation and eventually approach the snake. These coping models have been found to be superior to mastery or fearless models, who immediately display successful approach behavior very confidently and with no apprehension (Bandura, 1977a; Bandura, Adams & Beyer, 1977; Meichenbaum, 1971). The effective- ness of these coping models may have been enhanced by the combined effects of similar gender and similar fear between model and observer. Kazdin (1974), in a similar type of investigation, also found support pointing to the superiority of a coping model; however as Kazdin noted, the similarity dimensions of age and gender significantly enhanced and contributed to the overall effects. A series of investigations by Weinberg and his colleagues, although not viewed as modeling in the traditional sense, provide valu- able speculations concerning the experimental manipulation of model (confederate) similarity and dissimilarity (Weinberg, Gould & Jackson, 1979; Weinberg, Gould, Yukelson & Jackson, 1981; Weinberg, Yukelson & Jackson, 1980). Using a more motorically-oriented task, similarity was manipulated by having nonathletic subjects compete against a confederate presented as a varsity track athlete (dissimilarity) or against a con- federate presented as a nonathlete whose past history indicated a knee injury (similarity). This confederate did not serve as an instructional model; however,he did convey incidental information to the observer. Results revealed that subjects who were exposed to the nonathlete with a knee injury, extended their legs significantly longer than subjects exposed to the varsity track athlete. The results of these investiga- tions in the inhibition/disinhibition area of modeling including investigations in perseverance, offer support for Bandura's (1977a) contention that people's expectations of competence, or self-efficacy, which may be gained from their perceived similarity/dissimilarity to another individual in the same situation, influence how they behave, their choice of activities, how much effort they will expend, and how long they will persist in the face of obstacles and aversive experiences. One of the most pertinent studies regarding perceived similarity in task ability was conducted by Gould and Weiss (1981). One aspect of this investigation manipulated the similarity dimension by having non- athletic female subjects observe either a confederate female presented as a nonathlete (similarity), or a confederate male presented as a varsity track athlete (dissimilarity) perform a leg-endurance task. Since it has been posited by Bandura (1977a) that model similarity in- creases the personal relevance of the modeled performance to the observer, it was predicted that subjects perceiving themselves as similar to the model would persist significantly longer at the leg endurance task than would subjects viewing themselves as dissimilar to the model, resulting in less persistence on the experimental task. This hypothesis was sup- ported. Subjects observing a model of similar-gender, perceived to be similar in athletic ability, demonstrated greater leg endurance than sub- jects observing a dissimilar model of opposite gender perceived to be superior in athletic ability (Gould & Weiss, 1981). However, all of the investigations examining the inhibition/disinhibition area of modeling have never attempted to dissect which dimension of model similarity-- gender or perceived similarity in ability--is most salient to the observer. Although the modeling aspect of the Gould and Weiss investi- gation was one of the first studies specifically designed to manipulate perceived similarity in ability of a task related to the realm of phys- ical education, the similarity dimension was designed with the inclusion of model/observer gender similarity in order to maximize possible experi- mental effects. Unfortunately, this investigation could not determine which dimension of similarity, gender or perceived task ability, was the most salient to the observers. Most model similarity studies have been investigated within the conceptual framework of the disinhibition function of modeling. Thus, observers have generally been individuals who were trying to overcome a skills deficit. The use of similar models may also be an important technique for motivating already proficient performers to further improve upon their performance. However, this motivating function of modeling has not been investigated with competent observers. Consider for instance, the female actively engaged in sports tasks which have been viewed primarily as a male-oriented pursuit. Such women have typically overcome, or have learned to cope with, the social barriers and commonly held stereotypes concerning achievement-oriented women. It might be suggested that these women will not identify with the lower competency expectations for women, consciously or unconsciously held by society, but instead will perceive themselves as highly competent indi- viduals aspiring to higher levels of performance and ability strivings. It seems reasonable to suggest that women engaged in higher levels of athletics (i.e., intercollegiate competition) have experienced numerous reinforcing successful experiences; otherwise, they would have probably discontinued sports participation long ago and would not have aspired to such a high level of participation. Although purely speculative in nature, females categorized by the above description would probably perceive themselves as more similar in task ability to an athletic (male or female) model than a nonathletic model. In fact, these females nay perceive themselves as more similar in task ability to an athletic male model than an athletic female model due to the high achievement- orientation often identified with such woman. Athletic females may also be more motivated to perform by viewing an athletic male model who may provide more of a perceived challenge to them than an athletic female 9 model. Statement of the Problem The purpose of this investigation was to partially replicate and extend the modeling aspect of the 1981 study by Gould and Weiss. Specifically, the study investigated the differential effects of model similarity across two dimensions, gender and perceived task ability, on the performance of a leg-endurance task. A secondary aim was to investigate the effects of observers' athletic ability on modeling the leg-endurance task. Because Gould and Weiss (1981) also measured self- efficacy in their study, a third aim was to examine the effects of model similarity on self-efficacy. 10 Hypotheses The following hypotheses were investigated: 1. Female nonathletic subjects performed longest on the leg- endurance task when viewing the female nonathletic model. 2. Female athletic subjects performed longest on the leg- endurance task when viewing either the female or male athletic model. 3. Subjects viewing a model of similar athletic ability extended their legs longer than control subjects. No predictions were made concerning the most salient similarity cue--gender or perceived similarity in task ability--modeled by subjects due to a lack of previous knowledge concerning this issue. Delimitations This study was delimited to college-aged athletic and nonathletic female volunteers attending the Michigan State University. The differ- ential effects of model similarity on the performance of a leg-endurance task were investigated across two dimensions, gender and perceived task ability. Definitions The following operational definitions apply to the present investi- gation: Modeling--behavioral modifications resulting from an observer's exposure to another individual's performance or behavior (Bandura, 1969). Perceived similarity in task ability--either subjects viewing themselves as athletic and capable of performing up to the ability level 11 of the athletic model, or subjects viewing themselves as nonathletic and capable of a performance only comparable to that of the nonathletic model. Athletic experience--past or present athletic involvement, such as, membership on a high school athletic team, intercollegiate team, or any other organized sporting league excluding intramural participa- tion. Male or female athletic model--a male or female introduced to subjects as a current member of the Michigan State University Soccer Club team who lifts weights three times a week. Male or female nonathletic model--a male or female introduced to subjects as a fellow student with no previous athletic experience. Athletic female subjects--females currently participating (1982- 1983) on one of the Michigan State University varsity athletic teams or club athletic teams undergoing the same degree of rigor in practice and game schedules as that of the varsity athletic teams. Nonathletic female subjects--females indicating no organized* athletic experience, currently enrolled in the Michigan State University basic instructional physical activity program. Control_group--subjects performing the experimental task Without prior exposure to a model. Limitations A few limitations existed in this study. First of all, the number of female athletic subjects obtained was limited by the number of female varsity athletic teams at Michigan State University and by the number of female club athletic teams with practice and game schedules com- parable to that of the varsity athletic teams at Michigan State University. 12 This study was also limited by the fact that subjects were volunteers and were not randomly selected. Another limitation may have existed due to the gender of the experimenter. A female experimenter con- ducted this investigation which might have affected the results obtained, as well as possible experimental effects due to the exclusion of a male experimenter. An additional limitation of this study may have been imposed by the use of four different models. A final limitation of this study concerns the pre-experimental measurement of subjects' self-efficacy. A pre-performance measure of self-efficacy was obtained immediately before subjects engaged in the experimental task. For subjects exposed to the modeling conditions, this constituted a measurement of self-efficacy after having had observed the models. No measurement of self-efficacy was obtained before subjects were exposed to the models, making it impossible to determine initial self-efficacy changes in the subject as a result of observing the models. CHAPTER II REVIEW OF THE LITERATURE Modeling influences have been explained through associative theories (Allport, 1924), reinforcement theories (Gewirtz & Stingle, 1968) and affective feedback theories (Aronfreed, 1969). One of the most plausible and encompassing theories of modeling, however, is Bandura's (1969) stimulus-contiguity or social learning theory of modeling. Bandura's social learning theory includes informational as well as motivational explanations for the modeling process. The purpose of this chapter is to review social learning theory in terms of the informational and motivational components of the modeling process. This chapter presents a) the observational learning function of modeling, b) the response facilitation function of modeling, and c) the inhibition/disinhibition function of modeling on model/observer similarity. In its broadest sense, human behavior and performance can be viewed in terms of social learning theory (Bandura, 1969). The emphasis in social learning theory is on the vicarious, symbolic and self- regulatory processes involved in human psychological functioning. Bandura (1969) believes that human thought, affect and behavior can be influenced by observation (vicarious acquisition) as well as by direct experiences. In addition, symbolic processes are utilized by humankind 13 14 to represent events, analyze, plan, create and to engage in various other subsequent behaviors. According to Bandura, humans actively engage in numerous self-regulatory processes. People are active processors of their environments, selecting, organizing, and interpreting afferent stimuli (Bandura, 1969). Bandura acknowledges the self-directing capaci- ties of humans which produce self-generated incentives and outcomes allowing a certain degree of perceived individual control over their own lives (Bandura, 1977b). Social learning theory can also be applied to the modeling process. The concepts of vicarious, symbolic and self-regulatory psychological processes are evidenced in four basic components which Bandura (1969) has identified as essential to modeling theory. These four components are attention, retention, motoric capability and motivation. Bandura contends that for an observer to subsequently display the modeled act, he or she must first attend closely to and retain what was demonstrated. In addition, the observer must possess the motoric capability to repro- duce the modeled act and be motivated to do so. Of particular relevance to this investigation are the attentional subprocesses. The modeling process is greatly affected by observers' attentional styles and capacities, as well as a variety of other varia- bles both internal and external to the observers. Observers must attend to, recognize and discriminate between the distinctive features of the modeled act if effective reproduction is to be enhanced. Selec- tive attention is a key component of the attentional subprocesses, requiring the observer to know what must be attended to, how much of the act must be attentively observed, etc. For example, various model 15 characteristics may elicit differing amounts of selective attention from the observers. Characteristics such as competence and status of the model, and the degree of model/observer similarity along a number of dimensions such as gender and perceived similarity in task ability, may greatly influence the attentional subprocesses (Landers & Landers, 1973; Kazdin, 1974; Brown & Inouye, 1978). In order for an observer to subsequently display some aspect of the modeled act, the modeled response must be remembered by the observer. Cognitive processes play a vital role in the retention of modeled behaviors. According to Bandura (1969, 1971, 1977b), behavior patterns are coded and symbolically represented in memory. This symbolic representation can take two forms, imaginal and/or verbal representation. Once formed, these symbolic constructs are stored in the memory and are later called upon to guide future actions. Behavioral reproductions of modeled acts are achieved by the utilization of symbolic representations which aid in the guidance of~ the desired performances. Successful reprOduction is dependent upon the task, the amount of spatial and temporal organization required, as well as the observer's physical and developmental limitations and capabili- ties (Bandura, 1971). The motivational subprocesses of Bandura's (1969) theory are significantly applicable to this study. The motivational subprocesses of modeling theory are concerned with the factors needed to facilitate or motivate the observer toward attention, retention and reproduction of the modeled response. Although the conditions of the other subprocesses may be met, the observer may not reproduce the desired act unless 16 sufficiently motivated by at least one of the following motivational sources: extrinsic inventives, self-incentives, and/or vicariously based incentives. Once again model/observer characteristics such as status and competence of the model and dimensions of model/observer similarity reflect variables which may influence this motivational com- ponent. It must be realized that the full impact of_the modeling process will not be achieved if severe deficiencies are evidenced in any of the four modeling subprocesses. As proposed by Bandura (1969), each of the modeling subprocesses has an influence over the nature and degree of modeling. These sub- processes are evidenced in different degrees when considering the vari- ous functions of modeling. The following sections will discuss the observational learning, response facilitation and inhibition/disinhibi- tion functions of modeling as they relate to Bandura's four subprocesses of social learning theory. Observational Learning Function of Modeling The observational learning function of modeling is evidenced when models elicit novel behavior in the observer; behavior which has been previously absent from the observer's repertoire of responses. The observational learning function has been demonstrated through the modeling and subsequent performance of strategy-oriented tasks by the observer, such as the "shoot-the-moon" game and the Bachman ladder climbing task. Examples of observational learning are also commonly demonstrated through studies utilizing specially constructed novel modes of response. Bandura (1971) illustrates this phenomenon by referring to the unique, nonsensical words 'lickitstickit' and 17 'wetosmacko' and concludes that these words would not have been expressed by subjects had these expressions never been modeled. Much of the research examining this observational learning func- tion of modeling is related to the retentional and the attentional/ motivational subprocesses proposed by Bandura. An example of research related to the retention subprocesses of observational learning was con- ducted by Gerst (1971) who examined the role of symbolic coding in the learning and retention of the deaf manual language through the observa- tion of a model. Gerst employed a 4 x 2 (Symbolic Code x Time of Per- formance) experimental design. Specifically, college-aged male and female subjects having no knowledge of the deaf manual language, ob- served a filmed model perform a series of complex movements taken from this language. After observing the model, subjects spent 1 minute performing one of four activities. The first symbolic activity, summary labeling, involved the coding of the modeled language into concise items representing the actual shape of the manual movement. The second form of symbolic representation, imaginal coding, required the subjects to form a picture in their heads of the actual manual movement. The third form of symbolic coding required the subjects to provide a detailed verbal description of the movement, while the fourth group of subjects served as a control condition, engaged in a distraction task of counting the beats of a metranome. After engaging in one of these four activi- ties, subjects were asked to reproduce the modeled movements immedi- ately after the 1 minute coding or distraction activity and again after a period of 15 minutes. All three of Gerst's coding conditions facilitated motor reproduction of the manual language of the deaf as 18 compared to the poorer performances reproduced by the control subjects who did not engage in any form of symbolic or verbal coding. In addi- tion, all subjects performed better on the immediate test than on the delayed test; and subjects employing summary labeling reproduced approximately twice as many correct responses on the delayed (retention) test as compared to the reproduction capacities of the other groups. The major findings of the Gerst investigation offered support for the hypothesis that some form of symbolic coding serves an important func- tion in observational learning. The observational learning function of modeling has also been researched in relation to the attentional/motivational subprocesses. Numerous social factors and characteristics of the model have been found to exert influence on the attention and/or motivation of the observer. For example, it has been demonstrated that models similar to the observer in terms of interests (hobbies) and background (membership in a mutual organization) are attended to and imitated more often than models dis- similar in these qualities (Rosekrans, 1967). The attentional and motivational effects arising from the observa- tion of a model are further demonstrated when considering the status and competence of the model. Rosenbaum and Tucker (1962) noted a significant effect on adult imitative behavior when varying the competence of a model. Specifically, adult observers matched the behavior of a success- ful (competent) model more rapidly than that of an unsuccessful (incompe- tent) model in a series of experimental horse-racing games. Similar results were reported by Zimmerman and Blotner (1979) in an investigation with first and second grade children on the performance of a wire puzzle task. 19 Baron (1970) extended the Rosenbaum and Tucker (1962) studies with the addition of another independent variable--attraction toward the model. Attraction toward the model was induced by varying the degree of attitude similarity between the subjects and model. Results indicated an interactive effect between level of attraction toward the model and the model's competence. A high degree of attitude similarity between model and subject enhanced imitation when the model displayed high task competence (success), but hampered imitation when the model exhibited low task competence (unsuccessful model). The results of this investigation suggested that one variable influencing adult imita- tive behavior is the level of competence shown by a model; however, variations in level of attractiveness toward the model may have produced specific rather than generalized modeling effects (Baron, 1970). These investigations offer support for Bandura's (1969) contention that models who are considered experts, models demonstrating a high degree of compe- tence and skill, as well as high status models are more likely to elicit attention and act as more salient sources of potential observer behavior than models not possessing these qualities. Due to the interactive effects of numerous variables and the dif- fering degrees of modeling influences, it may be suggested that certain models or modeling conditions may also elicit a motivational as well as attentional effect upon observers. Landers and Landers (1973) provided elementary school children with either a highly skilled teacher, highly skilled peer, unskilled teacher, or unskilled peer as a model. Subjects watched the model perform on a Bachman ladder climbing task and were then asked to perform this same task. Results indicated that children 20 observing the skilled teacher performed better than children observing the unskilled teacher or the skilled peer. The attentional/motivational effects of an older, more highly skilled and higher status model appeared to be evident. However, children who had observed the unskilled peer performed better than children who had observed the unskilled teacher model. Perhaps this finding can be related to motivational conditions introduced when competing against a peer perceived as equal or lesser in ability, possibly mediated by the thought, "if he/she can do it, so can I". Hence, modeling may not only provide a means of relaying motor performance information, but may, in fact, play an equally significant role in observer attention/motivation. Response Facilitation Another function of modeling proposed by Bandura is that of response facilitation. Through response facilitation, the behaviors, actions and performance of others provide cues or prompts designed to stimulate the performance to one's own pre-existing responses (Bandura, 1971). Due to the facilitory nature and enhancing properties of this function, response facilitation can be classified under the motivational subprocesses outlined by Bandura (1969). Many examples of the response facilitation function of modeling offer a great degree of applicability to the sporting world. In addition to the example offered in Chapter I, Landers and Landers (Note 1) provide additional "real-life" examples of response facilitation, such as a young child exaggerating in his/her own style of play the qualities (stance, etc.) of a famous, idolized athlete. However, more laboratory-oriented and tightly controlled field 21 studies need to be conducted in order to determine the viability of this influence in sport. Inhibition/Disinhibition of Responses The inhibition/disinhibition function of modeling deals with performance decrements or increments as the result of observing a model's performance, as well as the subsequent consequences associated with that performance. This function of modeling has been the recipient of a great deal of attention; however, much of the inquiry and system- atic investigation of this modeling function has been conducted in disciplines other than physical education or athletics. The inhibition research in the area of modeling has been conducted primarily by Bandura and his associates (Bandura, 1973; Bandura & Walters, 1959). As noted by the above authors, examples of the inhibitory effects of modeling include observers who show decrements in a modeled class of behaviors as a result of observing the model's behavior produce punish- ing consequences. As noted by Bandura (1973), observing the aggressive acts of others being punished usually results in less imitative aggres- sion by the observers than seeing the aggressive behavior rewarded or unaccompanied by any obvious consequences. The disinhibitory function of modeling is evidenced when the observation of models engaging in threatening, persevering, or aversive behavior without negative consequences, results in the observer's in- creased performance of these formerly inhibited or discomforting be- haviors. Research of this nature has produced very interesting results when considering phobic reactions and avoidance behavior patterns 22 emitted by snake-phobic subjects (Bandura & Adams, 1977; Bandura & Barab, 1973; Bandura, Blanchard & Ritter, 1969; Kazdin, 1973, 1974; Kornhaber & Schroeder, 1975; Meichenbaum, 1971). Much of the research in this area suggests that c0ping models are the most effective means of reducing fear in phobic subjects. Coping models who are initially portrayed as displaying a similar degree of fear as the observer but then overcoming that fear and eventually approaching the feared object without harmful consequences elicit substantial approach behavior by the fearful observer. This type of coping model has been found to be superior to a mastery model who immediately displays approach behavior toward a phobic object with no apparent fear or hesitancy. For example, Meichenbaum (1971), employing 36 female, snake-phobic undergraduates, found coping models to be more effective in reducing subjects' fear than mastery models. In addition to the superiority of the coping models, models who engaged in self-verbalization were found to be more effective in reducing fear than were silent models. Bandura (1977a) reasons that coping and self-verbalization models are superior to mastery and silent models because they convey greater efficacy informa- tion to the observer by providing a strategy for accomplishing the task. Research in the disinhibition function of modeling as it is re- lated to the physical education domain has largely been concerned with perceived task competence, or self-efficacy, and the effects of participant-modeling. In participant-modeling, a model first demon- strates the task, then engages in the task together with the learner or offers physical guidance whenever necessary with the purpose of increas- ing successful experiences for the learner. The few investigations that 23 have been conducted in this area have employed tasks considered "risky", "unpleasant", possibly "dangerous", or in other words, "high-avoidance" tasks (Feltz, Landers & Raeder, 1979; Lewis, 1974; Weinberg, Sinardi & Jackson, 1982). All of this research has indicated that participant- modeling is superior to live or video-tape modeling conditions. For instance, as a result of participant-modeling, subjects of initially low perceived task competence have been found to increase in efficacy and task competence and subsequent participation in "high-avoidance" type physical education and athletic skills, such as found in gymnastics (Weinberg, Sinardi & Jackson, 1982), swimming (Lewis, 1974) and diving (Feltz, Landers & Raeder, 1979). When considering the disinhibition literature, it seems plausible that the disinhibition function of modeling is closely related to the attentional/motivational subprocesses proposed by Bandura. It can be suggested that the attentional/motivational subprocesses are highlighted throughout the disinhibition function of modeling by increasing the per- sonal relevance of the modeled performance to the observer (Bandura, 1969), thus perhaps increasing the observer's attention to the model and motivation to perform. The disinhibition function of modeling may enhance the attentional/motivational subprocesses by increasing the personal relevance of the modeled performance to the observer through a perceived condition of model/observer similarity. Regardless of the type of modeling function investigated (observa- tional learning, response facilitation, inhibition/disinhibition), the relationship between and effect of model/observer similarity has been the topic of much interest. Therefore, a more detailed discussion of the literature directly related to model/observer similarity is addressed in the following section. 24 Model/Observer Similarity Conditions of perceived similarity between model and observer can exist along a number of dimensions. Models may be similar to observers in age, gender, perceived task ability, fear, I.Q., etc. The examination of gender effects in the modeling literature has been concerned with the employment of models either of similar or dissimilar gender to the observer. However, the influence of similar or dissimilar gender models is far from clear. In his review on imitative behavior research, Flanders (1968) noted that some investigators (Bandura & Kupers, 1964; May, 1966; O'Connell, 1965) found that the gender of the model produced little or no main effects on observers' modeling behavior, while other investigators (Bandura & Huston, 1961; Bandura, Ross & Ross, 1963; Hetherington & Frankie, 1967; Hicks, 1965) have found interactive effects with regard to gender of the model and observer. For instance, Bandura & Kupers (1964) found that on a bowling task, 7- to 9-year-old boys and girls displayed patterns of self-reinforcement as the result of observing male and female models. Specifically, older, higher status models cfl’ either gender were imitated to a greater extent than were peer models of either gender. It was also determined that generosity of self-reward was in no way related to gender of the models or observers. On the other hand, Bandura et a1. (1963), when investigating the learn- ing and performance of aggression through modeling, found subject gender to be a highly significant factor, as well as some modeling effects attributable to model gender. Specifically, boys exhibited more total aggression than girls, and subjects observing a male model exhibited more aggressive "gun play" than subjects exposed to a female model. 25 In his review, Flanders (1968) also noted that most experiments usually exposed observers to the same model, or employed same gender, model/observer pairs without taking into account the gender of the model as a source of variance. Flanders concluded that, due to this frequent lack of concern regarding model gender as a significant source of variance, few dependable conclusions can be drawn when considering model/observer gender effects. According to Gould and Roberts (1982), models of the same gender as the observer will, at times, have a greater influence on the observer's motor performance than will models of the opposite gender. Many investigations have suggested that enhanced treatment effects are facilitated when same gender models are employed (Gould & Weiss, 1981; Kazdin, 1974; Perry & Perry, 1974; Rosekrans, 1967). However, in order to maximize treatment effects, much of the research in this area has employed multiple similarity variables making it difficult to differentiate the separate influences of the specific variables. In any case, most researchers seem to agree with the con— tention of social learning theory that model/observer characteristics such as similarity between model and observer will affect imitation in terms of the potential influence on the attentional and motivational subprocesses. Bandura (1971) noted that because similarity exists on numerous dimensions, interrelated factors can be attributed to the occurrence of many noted modeling effects. Similarity in terms of gender may not be the sole factor underlying the results. It is possible that other conditions of perceived similarity between model and observer combined with similarity along the dimension of gender may have pro- duced many of the reported findings (and vice versa). 26 For example, another source of similarity between model and observer, that of perceived task ability, has become an intriguing topic of concern. In this respect, the focus of much of the research into the disinhibition function of modeling has been concerned with a more abstract dimension of task ability, that of "approach" ability in fearful situations (Bandura & Adams, 1977; Borkovec, 1973). Approach ability, here, refers to the ability (or perceived task competence) of an individual to overcome fear exhibited toward an object or situation, approach that object and eventually handle or confront the situation. Research investigating this form of perceived task ability or task competence, has concluded that modeling can significantly help phobic individuals overcome their fear (Kazdin, 1973, 1974; Meichenbaum, 1971). Specifically, a condition of perceived similarity between model and observer in terms of initial fear towards a phobic object is a facili- tory factor greatly contributing to the success of treatment effects. However, it must be borne in mind that all of the phobic-behavior investi- gations, and the majority of modeling research, regardless of domain, has employed same gender models and observers. This notion was supported by Kazdin's (1973, 1974) investigations. Kazdin, in line with other disinhibition researchers, employed snake- phobic subjects, coping models and mastery models. In Kazdin's earliest study it was found that coping models led to greater avoidance reduction than mastery models. Similar results were found in Kazdin's 1974 study which also took into account similarity in terms of gender and age, besides similarity of initial fear of snakes. Kazdin noted that although beneficial effects provided by a caping model enhanced the reduction of 27 avoidance behavior, the inclusion of numerous dimensions of the model/ observer similarity variable (age, gender and amount of initial fear) may have significantly contributed to the overall effects. In any case, it can be suggested that a coping model, who may be perceived by the observer as similar to him/herself in terms of initial fear and avoidance behavior may raise efficacy expectations by instigating the, "if he/she can learn to overcome fear, so can I", motivational set of the observer. Meichenbaum (1971) concluded that the effectiveness of the coping model in reducing fear may be based on a) the condition of "perceived simi- larity between the observer and the model which facilitates imitation, that is the 'appropriateness' of the model for the observer" (Meichen- baum, 1971, p. 304) and/or b) demonstration of a specific strategy for approach-behavior which is conveyed to the observer as being an effec- tive means of learning to overcome fears. Similarity in perceived task ability has also been demonstrated in a number of investigations employing more "concrete" tasks, such as the word anagram task employed by Brown and Inouye (1978). Brown and Inouye (1978) randomly assigned 40 male college students to one of four social comparison situations. The first situation was referred to as a "similar competence" condition. In this condition subjects were led to believe that they were as equal in word-anagram competence as a male model whose failure on the word anagram task they had observed. A "higher competence" condition existed in which subjects were led to believe that they were superior in task competence to the unsuccessful male model. The third social comparison situation was a "no-feedback“ condition in which subjects were given no task ability information 28 regarding their own word anagram performance in relation to the perform- ance of the unsuccessful male model. Lastly, a control condition was established in which subjects worked on the experimental word anagram task without prior exposure to the male model. Following the above manipulations, subjects engaged in another series of word anagram tasks with the main dependent measure being the number of seconds the subjects persisted before giving up on the tasks. The results of this investigation led Brown and Inouye (1978) to confirm their hypothesis that a state of learned helplessness can, in fact, be induced through modeling, and that this condition of learned helplessness was mediated by the observers' perceptions of similarity to a model in terms of task competence. This learned helplessness phenomenon can be considered from a social learning theory perspective which purports that lowered performance expectations can be a consequence of direct failure or vicariously experienced failures (Bandura, 1977b). Vicariously induced helplessness occurs through the observation of a model who is perceived as similar in competence to the observer and continuously fails at a task. According to Brown and Inouye (1978), observing a successful model perceived as similar in ability to the observer would tend to create success expectations in that observer facilitating the observer's motivation to perform, and to perform well. On the other hand, observing someone similar in competence to oneself continually failing at a task would facilitate failure expectations in the observer, resulting in lack of motivation and persistence at the task. This is exactly what happened; subjects perceiving themselves as similar in ability to the 29 unsuccessful model persisted the least amount of time at the experi- mental word anagram task, while subjects perceiving themselves as more competent than the model persisted the longest when engaging in the word anagram task. A major difference between the Brown and Inouye (1978) study and other investigations employing a coping model perceived as similar (snake- avoidance studies) is that the model in the Brown and Inouye study never succeeded at the task, whereas all of the coping models in the phobic- behavior investigations eventually succeeded at the task. This observed difference may have influenced the observers' subsequent performance and beliefs regarding competence and the effects of perceived similarity in competence between model and observer since self-reflective thoughts may very well intensify an individual's efficacy beliefs by raising or lowering efforts concerning his or her task capabilities (Bandura, 1978). In the area of physical activities, the Lewis (1974) investigation is one of the few studies that employed coping models similar to the observer in initial task ability. in the investigation of participant- modeling. Lewis' study focused on the reduction of childrens' fear and avoidance of water and swimming activities. Forty elementary school-aged boys exhibiting fear of the water were assigned to one of four experi- mental conditions. These conditions consisted of a model plus partici- pation condition, a model only condition, a participation only condition and a control condition. The modeling plus participation treatment exposed subjects to a film of similar models in terms of age, race, gender and initial fear of the water. The models in the film displayed coping behavior by overcoming their initial fear of the water and 30 eventually, participating fearlessly in a variety of swimming activities. After viewing this film, subjects spent 10 minutes in the pool accompan- ied by the instructor. Subjects exposed to the model-only condition observed the same film but did not engage in subsequent water activi- ties. As a substitute, these subjects played a lO-minute game of pool- side checkers with the instructor. Participant-only subjects viewed an irrelevant film and then engaged in the lO-minute swimming activities with the instructor. Control subjects also viewed the irrelevant film and then participated in the lO-minute session of poolside checkers. Results of this investigation indicated that the modeling plus participation condition resulted in the most anxiety reduction and increase in swimming behavior, followed by the participation only treat- ment and then the modeling only treatment. Lastly, all of the treatment conditions were more effective than the control condition. Lewis' (1974) investigation supports the effectiveness of participant-modeling and the general consensus of similarity research which suggests that model/observer similarity along a number of dimensions produces the greatest modeling effects. Although designed for the purposes of self-efficacy research and not for the specific investigation of model similarity, a series of investigations by Weinberg and his colleagues provide pertinent informa- tion regarding the experimental manipulation of perceived similarity/ dissimilarity between model and observer in terms of task ability or task competence (Weinberg et al., 1979; Weinberg et al., 1981; Weinberg et al., 1980). The way in which self-efficacy was manipulated in Weinberg's studies was by having nonathletic subjects compete against a 31 confederate presented as a varsity track athlete (dissimilarity in task ability) or against a confederate presented as a nonathlete whose past history acknowledged a knee injury (similarity in task ability). Results revealed that subjects exposed to the nonathlete with a knee injury extended their legs significantly longer than subjects exposed to the varsity track athlete. In other words, perceived similarity along the dimension of task ability (competence) seemed to have enhanced the observers' subsequent performance, while perceived dissimilarity between model and observer in task ability had a negative effect upon observers' performance. However, it must be remembered that other variables may have influenced the experimental results, as the purpose of these investigations was not to isolate the amount of influence exerted by various dimensions of similarity cues. One investigation employing the same task as Weinberg and his colleagues, was designed by Gould and Weiss (1981) for the purpose of examining the effects of model/observer similarity. Specifically, this study was designed to determine whether the observation of a similar or dissimilar model who emitted varying self-efficacy statements (self- talk) influenced the observers' efficacy expectations and motor perform- ance. Gould and Weiss contended that efficacy appraisals may be greatly influenced by perceived similarities between observer and model on characteristics such as gender, motor skill level, etc., thus affecting the observers' subsequent performance (Gould & Weiss, 1981). One hundred and fifty nonathletic, college female volunteers enrolled in elective physical activity courses, served as subjects for this investigation. The design consisted of a Model Similarity x Model 32 Talk x Trials factorial design and a No Model Control condition. Model/observer similarity was induced by having the nonathletic females observe either a confederate female presented as a nonathlete (similarity), or a confederate male presented as a varsity track athlete (dissimilarity). The four conditions of model talk consisted of a) a positive self-talk model who performed and emitted positive self-efficacy statements, b) a negative self-talk model uttering negative self- confidence statements, c) an irrelavant-talk model who performed and spoke of items unrelated to the task, and d) a no-talk model who per- formed in silence. Pre-and post-self-efficacy questionnaires assessed the subjects' levels and strength of self-efficacy concerning the task. Results of this investigation revealed that subjects exposed to the model who was similar in task ability and gender (the nonathletic female), extended their legs significantly longer than subjects exposed to the dissimilar model (the athletic male). Results of the exposure to the model talk-no talk conditions revealed that the similar-positive and similar-no talk groups extended their legs significantly longer than the dissimilar-positive or negative talk groups and also longer than the no-model control group. However, subjects' self-efficacy levels were not found to be the major variable affecting performance changes. Although the modeling aspect of the 1981 Gould and Weiss investi- gation was one of the first studies specifically designed to manipulate perceived similarity in ability of a task related to sport, the simi- larity dimension was designed with the inclusion of model/observer gender similarity in order to maximize experimental effects. As a result, this investigation could not determine which dimension of 33 similarity, gender or perceived task ability, was the most salient to the observers. As demonstrated by previous studies (Weinberg et al., 1979; Weinberg et al., 1981; Weinberg et al., 1980; Gould & Weiss, 1981), the influencing role of personal cognitions must be considered when examining human behavior studies. Bandura (1977a) provides a theory of self-efficacy which addresses the issue of how cognitions may mediate performance. Bandura (1977a) perceives self-efficacy as a situationally specific construct influencing one's particular choice of activities and the degree of effort and persistence generated when dealing with a certain task. Bandura (1977a) views modeling as providing a medium through which personal expectations and feelings of self- efficacy can be influenced. Vicarious experiences can alter expecta- tions. For example, within the inhibition/disinhibition modeling func- tion, observing others engaging in threatening behavior unaccompanied by observable adversive consequences can generate persistence efforts and alter expectations among observers that they, also, can be success- ful at the same activity, perhaps mediated by the notion, “if he/she can do it, so can 1". Modeling can affect expectations of personal efficacy, which in turn can affect subsequent behavior. Increasing the personal relevance of the modeled act to the performer (e.g., by manipulating model/ observer similarity) can alter observers' perceptions of their own per- formance capabilities. For instance, modeling techniques such as those techniques successfully used in reducing avoidance behavior, may modify behavior through the mediating influence of efficacy expectations. 34 In conclusion, the mediating role that cognitions such as self- efficacy exert on performance must not be overlooked; modeling is a process which influences self-efficacy (and expectations) which in turn intervenes to influence performance. Perceived similaritygand proficient performers. The majority of model similarity studies have examined variables that influence the disinhibition function of modeling, dealing with observers possessing low expectations concerning their capabilities. In sport tasks, which typically have been viewed as male-oriented, many women do perceive themselves as low in ability. When dealing with research of this nature the socialization process must be considered. According to the conflict-enculturation model (Roberts 8 Sutton- Smith, 1962), every social system must consider the 'cultural maintenance' problems encountered which must be resolved in order for the system to survive as an intact and continuing entity (Birrell, 1978). One remedy utilized by the system is the employment of socialization practices. These are used to "ease individuals into necessary roles or perspectives" (Birrell, 1978, p. 63). As a result, the by-products of these socializa- tion practices include commonly held gender—role stereotypes and myths, attribution, achievement-motivation and success-failure theories, which have attempted to offer plausible explanations concerning the lower levels of strivings, competence and achievements noted among a large number of females (Deaux & Ferris, 1977; Feldman-Summers & Kiesler, 1974; House, 1974; Kidd & Woodman, 1975; Lenney, Browning & Mitchell, 1980; McHugh, Duquin & Frieze, 1978). However, the literature has recently noted that not all women are underachievers who hold low expectations of their 35 capabilities (Lenney, 1977; Ogilvie, 1978). This has become increasingly evident as represented by the upsurge of women holding positions of high status, as well as the increase of women entering the upper echelon of the sporting world. The use of similar models may also provide important cues and motivational properties for further enhancing the performance of these already proficient performers. This function of modeling has not been investigated with competent observers. For example, competent female athletes have very rarely been provided with similar role models. It might be suggested that these women will not identify with the lower female competence expectations held by society, but will perceive them- selves as highly competent and will aspire to higher levels of perform- ance. Females in higher levels of athletics probably have had numerous successful and reinforcing experiences. These females are experiencing higher and higher levels of personal status and competence. Such women probably perceive themselves as more similar in task ability to a competent model (athletic) than to an incompetent (nonathletic) model. Although purely speculative in nature, some female athletes may perceive themselves as more similar in task ability to a male athlete than to a female athletic model due to the high achievement strivings typifying many women actively engaged in athletic competition. A male athletic model may in fact, provide many athletic women with more of a perceived challenge than would the observation of a female athlete, resulting in a higher amount of motivation to out-perform a male athlete. Of course, the saliency of the similarity cues provided--gender and perceived task 36 competence--will most probably determine the amount and type of modeling effects encountered. Thus, in view of the complexity and controversy of the mitigating effects of model/observer similarity along the dimensions of gender and perceived task ability, the following study was undertaken to examine the saliency of these two dimensions of the similarity domain. In addi- tion, this investigation was conducted to determine the most salient similarity modeling cue necessary for different populations of observers (e.g., athletic and nonathletic observers), as well as the inclusion of more research on female athletes. CHAPTER III METHODS AND PROCEDURES The inhibitory/disinhibitory function of modeling was investigated by studying the differential effects of model similarity upon motor per- formance of a leg-endurance task across two dimensions, gender and perceived task ability. Because previous investigations were designed for the purpose of maximizing experimental outcomes rather than dissect- ing the similarity dimension, the question still remains as to which dimension of model similarity--gender or perceived similarity in ability-- is most salient to the observer. This investigation was conducted to answer the preceding question in hopes of contributing more information to the current store of knowledge concerning this issue. Since the self- talk dimension of the Gould and Weiss (1981) investigation revealed that the similar positive-talk modeling condition elicited the same degree of beneficial effects as the similar no-talk modeling condition, the self- talk dimension was eliminated from this investigation. Also, in addition to the inclusion of a male nonathletic model and a female athletic model, the present replication and extension of the modeling aspect of the Gould and Weiss (1981) study investigated the effects of observers' athletic ability on modeling the leg-endurance task. 37 38 Subject and Design The subjects were 150 female students enrolled at the Michigan State University. Subjects were composed of athletic volunteers cur- rently participating (1982-1983) on one of the Michigan State University varsity athletic or club athletic teams. The varsity athletic teams represented in this study included the volleyball, softball, basketball, gymnastic, swimming and tennis teams. Female athletes who partici- pated on the Michigan State University Rugby Club team were chosen as potential subjects due to the rigorous practice and game schedule of this team which is comparable to the game and practice schedules of female varsity athletic teams at Michigan State University. The only reason the Rugby Club team has not achieved varsity status is that they do not possess the necessary financial backing needed to qualify as a varsity team. Subjects were also composed of nonathletic volunteers enrolled in the Michigan State University elective physical activity courses. Only volunteers indicating no physical limitations which might prevent them in any way from reproducing the modeled demonstration were selected as subjects. In addition, only volunteers indicating no previ- ous contact with or knowledge of the experimenter and her background were selected as subjects. The study employed a 2 x 2 x 2 factorial design, with the first factor being the gender of the models: males versus females; the second factor being the athletic ability of the models: athletic models and nonathletic models; and the third factor being the athletic ability of the observers: athletic and nonathletic observers. A table of random numbers was used to assign each of the 75 athletic and 75 nonathletic subjects to one of the four treatment conditions, or a no-model control, 39 with the resrriction that there be no more than 15 athletic and 15 nonathletic subjects per condition. Experimental Task Subjects were asked to perform a modified version of a muscular endurance task (Gould & Weiss, 1981) which required the subject to extend her dominant leg in a horizontal position, maintaining that posi- tion for as long as possible (see Figure l). The subject sat on an adjustable swivel stool, thus allowing a lower leg-length correction to be made prior to performance. This was done to avoid differing angles of the subject's non-lifting leg. The subject folded her arms across her chest and sat upright with her back against the wall. The shoeless dominant leg was extended above and across a white cord that was sus- pended by a wooden apparatus specifically designed for this investiga- tion. The height of the cord was adjusted to each subject so that it was equal to the height of the stool. The subject's performance was timed, using a Singer Industrial Timer Corporation electronic stopclock (model SC-lOO). In preparation for the task, the experimenter supported the subject's extended leg in a horizontal position above the cord. When the subject was ready to begin the task, the subject lifted her leg off of the experimenter's supporting hand. As soon as the subject's leg was no longer in contact with the experimenter's hand, the clock was started. When the subject could no longer hold the extended leg above the cord, the leg lowered, contacting the cord, automatically stopping the timer. As a precautionary measure, the adjustable stool which had wheels on the bottom of all four legs, was placed on a piece of matting to prevent any possible slipping or rolling of the stool. 41 Model Similaritleissimilarity Manipulations A 60-second video-taped demonstration of one of the four models performing the experimental task was employed in this investigation. A video-taped demonstration was used instead of a live demonstration in order to control for consistency of model performance; furthermore, research has shown that televised models are as effective as live models on observers' imitative behavior (Bandura & Menlove, 1968; Bandura, Ross & Ross, 1963; Klinger, 1967). All models were filmed using an Hitachi (model VK-C 1000) color video camera. Each model's performance was recorded on a separate Sony Dynamicron L-250 color video-cassette. Each model was filmed extending his/her right leg, and a 3/4 view of the model was taken from the model's right side. Subjects randomly assigned to the modeling conditions observed one of the follow- ing video-taped demonstrations: l. A nonathletic female model--who was introduced as a fellow student with no previous athletic experience. 2. An athletic female model--who was introduced to subjects as a current member of the Michigan State University Soccer Club team who also lifted weights three times a week. 3. A nonathletic male model--who was introduced as a fellow student with no previous athletic experience. 4. An athletic male model--who was introduced to subjects as a current member of the Michigan State University Soccer Club team who also lifted weights three times a week. Athletic or nonathletic introductions of the models were provided in order to manipulate and enhance a condition of similarity/dissimilarity 42 along the dimension of perceived task ability between the subject and the model. In addition, the athletic manipulations were achieved by selecting two models (male and female) each of whom possessed well- defined musculatures and healthy, athletic-looking physiques. Both models were matched on personal characteristics as much as possible. The two models were both sport psychology graduate students and former athletes with similar physical appearances. They were also both brunettes, approximately the same age, height, build and coloring. They both performed in running shorts and emitted similar nonverbal gestures. The two nonathletic models were matched on personal character- istics as much as possible. Both were nonathletic graduate students, somewhat slight in stature and thin. The two nonathletic models were both the same age, build and coloring, with very undefined musculature. They performed in longer-length, baggy shorts and emitted similar non- verbal gestures. Lastly, all four models were depicted as successful on the leg-endurance task; that is, none of the video-taped films showed a model lowering his/her leg onto the cord, stopping the clock. Models were depicted as successful in order to eliminate any confounding effect due to the observation of an unsuccessful model. The use of only two models, one female portrayed as both athletic and nonathletic, and one male portrayed as both athletic and nonathletic, may have decreased the number of possible confounding variables due to the individual characteristics of each of the models. However, the decision to use four models was made in order to maximize the experi- mental athletic/nonathletic manipulations by providing models of differ— ing body types instead of relying solely on verbal descriptions of the models' athletic (or nonathletic) backgrounds. To check whether or not 43 model body type was one of the main characteristics that differentiated one model from another, or whether the models were perceived as differ- ing on numerous other characteristics, fourteen members of an under- graduate physical education majors class were shown the video-tapes of all four models performing the experimental task and asked to rate them on the following 10 characteristics: self-esteem, intellectual ability, attractiveness, athletic ability, self-confidence, autonomy ' (independence), emotional control, muscular endurance, outgoing person- ality and affiliation (group—oriented). Students rated the models on a 7-point scale ranging from "below average (low)" to "above average (high)" on each one of the 10 characteristics. In order to determine which of the 10 characteristics differentiated the four models, students' responses were submitted to a discriminant function analysis. Results indicated that only the first discriminant coefficient was significant, x2 (21) = 125.20, p_< .00001. The char- acteristics that discriminated the four models the most were athletic ability and muscular endurance. These results supported the athletic/ nonathletic manipulation of models. However, it should be noted that the subject to variable ratio was not very high. The discriminant func- tion coefficients as well as the means and standard deviations are con- tained in Table 1 (on the following page). Manipulation Check A manipulation check was used to determine the effectiveness of the similarity/dissimilarity manipulations. This was assessed by ask- ing the subjects to respond to the question, "How similar do you perceive yourself to be to the person who demonstrated this task?" on a 7-point scale ranging from very dissimilar to very similar. 44 .mwmapmcm comuuczw ucmcmswcumwn Low emcmucm poc .mgommsmgu .mcmz new emu» mococmpou msp cmpwmw movamwcmpumcmgu mmcgke Aumucmwco . ANN.V no.” Amm.v ¢_.¢ Amov m¢.m Ape.v am.¢ -a:oemv compawppcc< . A-.V mm.m “mm.v 4F.e Amo.v me.» “_e.v m~.¢ achaeomzma mgwomoso mom. Ame.v m¢.~ Amm.v P~.m hem.v ep.N Aom.v 48.8 augmezucu Empaumaz * ANN.V ma.m Amm.v e_.¢ Amo.v me.m Apm.v m~.¢ Poeoeoo Pm=o_poem _op. A-.V mm.m Amm.v e..¢ Amm.v me.m Aem.v a~.¢ Aaucmccmaaeewv ssocop=< om~.- Ame.v mk.m Amm.v PN.¢ Apm.v m~.m Amo.v om.¢ mucouwccou-c_mm Fae. Ame.v em.~ “No.v ma.m Amm.v no.N A~¢.V -.m seepea< owampgp< ems. ANN.V mm.m Amm.v ¢_.e Amm.v m~.m APG.V a~.¢ mmmcm>epuaeou< ewe. Amm.v oo.¢ Amm.v ep.e Am~.v km.m A_o.v m~.¢ spm_ma< Paaoume_ap=H APO. ANN.V mm.m Amm.v FN.e Ae~.v mm.m Aem.v om.¢ emaama-eeom mm. m. _mw mm mm. “a mm. “a mo=o_upccaoo Paco: oeum_;om pave: Paco: oepapgpa Fave: moeomcempuazmgu Page: umNCuemucmpm -coz aFMEJE ucoa_;o< dengue -coz ape: ueaang< aFaz waxy Pena: an muwummcwuumcmgu Paco: cm» on» com mucmmowmmmoo cowpucsu acacwemgumwo ecu mcowumm>wo ccmucmum .mcmwz p mpnmh 45 Questionnaires Pre-experimental and post-experimental questionnaires were dis- tributed to all subjects. In addition to determining the effectiveness of the similarity/dissimilarity manipulation, these questionnaires assessed level and strength of efficacy expectations as well as assess- ing subjects' perceptions of competition with the model, subjects' sense of model influence and subjects' perceptions of their own perform- ance in relation to the models' performance. The test/retest relia- bility of the self-efficacy questionnaire items, determined by another subject population, ranged from r = .81 to r = .99. Pre-experimental questionnaire. The pre-experimental question- naire, administered just before attempting the task, consisted of three background information items as well as 18 items designed to assess level and strength of efficacy expectations. The background items and the first three self-efficacy questions were answered by checking the appropriate response and by circling the number best reflecting the subject's feelings on a likert-type scale. The 15 remaining self- efficacy items consisted of 15 specific time designations ranging from 30 seconds to 4 minutes; at l5-second intervals. Subjects were required to place a check mark next to each time designation that they felt capable of performing as well as checking their percent certainty of performance for each time designation checked. The actual question- naire is contained in Appendix A. Post-experimental questionnaire. The post-experimental question- naire, administered to subjects upon the completion of the leg-lift task, was designed to assess level and strength of efficacy expectations 46 for future performance of the experimental task. The first self-efficacy question was answered by checking the appropriate response and by circling the number best reflecting the subject's feelings on a likert-type scale. The 15 remaining efficacy items consisted of the same time designations as found on the pre-experimental questionnaire and were answered by subjects in the same manner. Also, subjects were asked to respond by placing a check mark on the appropriate line indi- cating who they thought would perform best on the experimental task, males or females. Four additional questions (two open-ended items) appeared on the questionnaires of the subjects exposed to the modeling conditions. These questions dealt with the subject's perception of similarity to the model, the subject's perception of her performance in relation to the model's performance, as well as assessing the subject's sense of competition with the model. The actual questionnaire is con- tained in Appendix 8. Pilot Study A pilot study was conducted prior to the actual testing of the experimental subjects. Twelve Michigan State University female graduate students (six athletic and six nonathletic) were randomly assigned to one of the modeling treatment conditions. The pilot subjects performed the experimental task, simulating actual experimental procedures and conditions. The purpose of the pilot study was to determine the subjects' understanding of the instructions and questionnaires, the effectiveness of the similarity/dissimilarity manipulation, performance variations in the amount of time subjects maintained their legs in the extended posi- tion, and the amount of time necessary for each subject to complete the 47 entire experimental procedure. The pilot study was also conducted as a practice run for the experimenter to ensure smooth administration of actual testing procedures, as well as identifying any necessary pro- cedural adjustments. Results obtained from the pilot study indicated that all subjects understood experimental instructions and questionnaires. The similarity by athletic ability manipulation was effective, in that all athletic subjects exposed to an athletic model rated themselves as "somewhat" to "very similar" to that model and all nonathletic subjects exposed to a nonathletic model rated themselves as "somewhat" to "very similar" to the nonathletic model. Regarding the dissimilarity athletic manipula- tion, subjects exposed to a dissimilar model in terms of athletic ability did in fact indicate that they perceived themselves to be very dissimilar to the observed model. In addition it was determined from the pilot study that 25 to 30 minutes were necessary for each subject to complete the entire experimental procedure, and that performance variations in the amount of time subjects maintained their legs in the extended position ranged from 34 seconds to 4 minutes and 15 seconds. Lastly, it was concluded that no procedural adjustments were necessary before the initiation of the testing of actual experimental subjects. Procedure Testing environment. The subjects were individually tested in the sport psychology laboratory at the Michigan State University. The laboratory contained a table and pencils for the completion of the informed consents and questionnaires, a NEC Auto color television, a Sony Betamax portable videocassette recorder and a Sony AC power adaptor 48 for viewing the modeled performance, the experimental task (adjustable stool and leg-lift apparatus with clock) and the experimenter's data collection sheets (see Figure 2). The number of trials (three) and the length of rest periods were determined from the pilot study and from previous research (Gould & Weiss, 1981). A 60-second rest period was selected by Gould and Weiss in the 1981 investigation. Gould and Weiss selected 60-seconds because it was not an adequate recovery time period between trials and felt that the motivational component provided by the model would be maximized when the subject experienced an unpleasant and adversive performance environment. The 60-second rest period between trials was timed using a stopwatch. Administration of test. Upon arrival at the testing room, the subject was greeted by the experimenter and given a brief verbal explana- tion of the experimental task. The following explanation was read to the subjects: This is a simple muscular endurance task, much like situps or pushups, designed to assess the muscular endurance of college students throughout the United States. You are to sit on this stool, fold your arms across your chest and keep your back flat against the wall. Please make sure you keep the foot of your nondominant leg flat on the floor. You are to extend your dominant leg above and across the white cord and hold your leg above the cord for as long as possible. Initially, I will sup- port your extended leg with my hand. When you are ready, you will lift your leg off of my supporting hand. As soon as your leg is no longer in contact with my hand, I will start the clock. When you can no longer hold your leg in the extended position above the cord, your leg will lower, contacting the cord and stopping the clock. You will have three trials with a 60-second rest between trials. Remember, you are to keep your dominant leg extended for as long as possible. After listening to the explanation, the experimenter clarified any questions; the subject completed an informed consent and the experimenter . . . I . . . s . . Tl _ t . . 1| . Ck . . . :1. es fl . h... .36 8 ,. . 1 b0 1 . . v ql. u .I. g . Ma HI. . S . . . . . 1111111111 . - - """" N ''''''' fl "' N * - . - . . . . . . 0 . . . .: _ t . _ . """" N°"'l - p er . . . a ter . . . . ddtdek . . . . anerds . u ._.._.. . _ asone n 5 . . W C? 3 . . . o . . . . S . P . . IL . u U — . . 0 .1. T 1 T .1. . . 0 4 m .l. A 3 . . S .1 2 A 3 . . . . P . . . . . P . . . . . . . . . . . . . . . . u . . . 1| _ . I . . .I. . . . . . . 1 Testing room. Figure 2. 50 obtained and recorded the subject's background and information (i.e., age, year in school, major). A subject in the control condition received a brief explanation of the pre-experimental questionnaire directions, completed the pre— experimental questionnaire, and was then asked to remove her shoes in order to eliminate any variations between subjects due to shoe heel- height of the nonlifting leg, and to eliminate any excessive weight on the lifting leg. The swivel stool was adjusted to each subject so that the lower leg formed a right angle with the upper leg, and the bottom of the nonlifting leg rested flat on the floor. The control subject was then asked to fold her arms across her chest, reminded to keep the foot of her nondominant leg flat on the floor and her back flat against the wall. The control subject extended her dominant leg above and across the white cord while the experimenter supported the extended leg. The control subject was instructed to begin whenever she felt ready. Upon completion of the three trials, the control subject was given a brief explanation of the post-task questionnaire directions and then given the post-task questionnaire to complete. Upon completion of the questionnaire, the subject was told that she had done a fine job, was thanked for her participation and was asked not to discuss the experiment with anyone. After the experimental subject listened to the task explanation and received any clarifications of questions, completed the informed consent and provided the experimenter with background information, the experimenter read the following statement: In order to show you how the task is to be done, I want you to carefully observe a videotaped demonstration of another person who was asked to perform this task. Please watch closely how 51 this person positions himself (herself) and the way in which he (she) approaches this task. Then depending upon the subject's assigned modeling condition, one of the following statements was presented: 1. The woman (man) you will see performing is a fellow student at Michigan State University with no previous athletic experience. She (he) has never participated on a high school athletic team, intercollegiate team, or any other organized sporting league, excluding intramural participation. 2. The woman (man) you will see performing is a current member of the Michigan State University Soccer Club team who also lifts weights three times a week. After the experimental subject was provided with one of the above statements, the appropriate video-tape was turned on and the experimental subject observed the performing model. The video-tape was then turned off and the subject was asked if she knew the model she had observed. Subjects indicating no previous knowledge of the model received a brief explanation of the pre-experimental questionnaire directions and then proceeded to the pre-experimental questionnaire. Any subject having prior knowledge of the model was thanked for her help and dismissed with the explanation that only subjects having no knowledge of the model could be used in this experiment. The experimental subject with no knowledge of the model was asked to remove her shoes and the necessary stool adjust- ment was made. The experimental subject was then asked to fold her arms across her chest, reminded to keep the foot of her nondominant leg flat on the floor and her back flat against the wall. The subject then extended her dominant leg above and across the cord, received the same amount of leg support as did the control subject, and completed the experimental procedure in the same manner as the control subject. 52 Upon completion of the entire experiment (data analyses and conclusions) all subjects were fully debriefed by means of a brief summary report which was mailed to each one of the subjects. Treatment of Data Motor performance scores for the experimental groups were analyzed within a 2 x 2 x 2 x 3 (Model Gender x Model Athletic Ability x Subject Athletic Ability X Trials) analysis of variance (ANOVA) with repeated measures on the last factor. In the event of significant differences for the interactions, post hoc tests were performed by means of the Tukey WSD procedure. A separate ANOVA was conducted to compare the experimental subject modeling conditions to the athletic and nonathletic subject control groups. This was accomplished by means of a 2 x 2 x 3 (Experimental/ Control Groups x Subject Athletic Ability x Trials) ANOVA. Similarly, post hoc tests were performed using the Tukey WSD procedure for equal n comparisons and Behren's Fisher 3} tests (Kohr, 1970) for unequal n comparisons, in the event of significant differences for the interactions. In order to directly test the hypotheses stated in Chapter I, a priori contrasts were conducted for each hypothesis. Questionnaire items concerning self-efficacy as well as the ques- tionnaire item asking experimental subjects to rate their perceived similarity to the observed model, were analyzed within separate ANOVAs. Specifically, for the experimental groups, self-efficacy was analyzed within a 2 x 2 x 2 (Model Gender x Model Athletic Ability x Subject Athletic Ability) ANOVA or within a 2 x 2 x 2 x 2 (Model Gender x Model Athletic Ability x Subject Athletic Ability x Pre/Post Scores) ANOVA. 53 To compare the experimental subject modeling conditions to the athletic and nonathletic control groups, self-efficacy was analyzed with a 2 x 2 (Experimental/Control Groups x Subject Athletic Ability) ANOVA or within a 2 x 2 x 2 (Experimental/Control Groups x Subject Athletic Ability xPre/Post Scores) ANOVA. Perceived similarity for the experi- mental-only groups was analyzed by means of a 2 x 2 x 2 (Model Gender x Model Athletic Ability x Subject Athletic Ability) ANOVA. Post hoc tests were performed (Tukey WSD or Behren's Fisher 3} tests for unequal nfs) in the event of significant differences for the interactions. Lastly, descriptive statistics (frequency distributions and percentages) by means of a crosstabs analysis were performed for all other question- naire and background information items. CHAPTER IV RESULTS Seventy-five athletic and 75 nonathletic college-aged females were selected for this study. Subjects were randomly assigned to one of four modeling conditions or were assigned to a no-model control condi- tion to investigate the influence of two dimensions of model/observer similarity cues (gender and perceived task ability) on the motor perform- ance and self-efficacy of a leg—endurance task. The effects of observers' athletic ability on modeling the experimental task and self-efficacy was also examined. A 2 x 2 x 2 x 3 (Model Gender x Model Athletic Ability x Subject Athletic Ability x Trials) factorial design plus an external control group for each subject ability group was employed. The task required subjects to sit on a stool and extend their legs for as long as possible above and across a cord for three trials with a one-minute rest between trials. The length of time the leg remained extended above the cord was recorded. Pre- and post-experimental questionnaires were administered to all subjects to determine the effectiveness of the model similarity/dissimilarity manipulation, level and strength of subject self-efficacy, subjects' perceptions of competition with the model and subjects' perceptions of their own performance in relation to the model's performance. 54 55 The results of this study have been organized into two sections. The first section presents the motor performance results. The second section is concerned with subjects' questionnaire data. All of the results in this chapter are reported at the .05 level of significance. Motor Performance Results Motor performance scores on the leg-endurance task for the experi- mental-only groups were analyzed within a 2 x 2 x 2 x 3 (Model Gender x Model Athletic Ability x Subject Athletic Ability x Trials) ANOVA with repeated measures on the last factor. The means and standard devia- tions for motor performance in each condition plus the control group on all three trials are reported in Table 2. Results from this analysis indicated that the main effect for subject athletic ability was signifi- cant, E (l, 112) = 39.87, p_< .0009. This main effect indicated that athletic subjects extended their legs significantly longer (M.= 101.25, ‘§Q_= 46.54) than nonathletic subjects (M = 59.03, §Q_= 23.57). In addi- tion, there was a trend toward a significant interaction (p_>’.059) between Model Gender and Model Athletic Ability. Subjects observing a female athletic model tended to extend their legs longer (M = 89.17, SD 46.12) than subjects observing a male athletic model (M = 72.76, .SD 24.83). This trend can be seen in Figure 3. In addition, this trend toward an interaction is probably better interpreted by examining the trend (p_> .087) toward the Model Gender by Model Athletic Ability by Subject Athletic Ability interaction. The two-way trend toward an interaction was mainly due to the motor performance of athletic subjects. Athletic subjects observing a female athletic model tended to extend their legs longer (M_= 115.16, SD = 39.52) than athletic subjects 56 Aem.¢¢v o¢.P~ ASN.4.V oo.~¢ Akm.P~V -.Ne Amm.mmv m~.¢m Amm.mwv m~.mk Awe..mv o~.~m Apm.emv mm.mm Aem.~_v Am.~m Amm.~mv mm.Ne Aeo.mmv mm.mm Apo.2~v mm.Nm Amp.mmv om.moP Aeo.omv m~.am “_m.va Ne.om Ame.kev me.mk Amm.mmv -.mP_ Amm.Pev mk.oup A¢_.okv N¢.k¢~ Amm.mpv mp.~m Awk.m~v om.mm Amo.w_v mm.¢m Anm.~ev o~.mo_ Amm.omv em._o Aeo.mmv mm.~m, Amm.mwv mk.mm Amo.F~v oe.- flop.-v om.km “Pm.omv oo.¢~ flop.mmv o¢.Nk Ako.FmV om.mo_ mpuahnzm acumpgoacoz muuannam u_pmpgp< m Fawze muumnnsm owomegpmcoz muumnasm uepapga< N _awze muumhnam awoapsoacoz muomhnzm uwpm_;p< F _awcc am m. cm z cm x _meo: _auoz uwompgpacoz uepmrxpa mpocucou mpmuoz wmemu am 2 am 2 Page: Paco: uwompsuacoz u_om_;o< “Page: ape: muumnnzm pocucou mzmcm> paucwewcmaxm eo mucmecoecma Logo: cow m=o_umw>mo ucmucaum new memo: N m_nae 57 O 0 female Mun Performance 111110!ch 7O Athletic Model Nonathletic Model MODEL ATH LETIC ABILITY Figure 3. Model Gender by Model Athletic Ability Interaction on performance. 58 observing a male athletic model (M_= 84.20, SQ,= 23.20). There was very little difference in motor performance between the four modeling conditions for the nonathletic subjects (see Figure 4). The Trials main effect was significant (p_< .0009), E = 49.78, df = 1,158]. Post hoc analysis using a Tukey WSD procedure (Winer, 1971), indicated that performance times significantly decreased from Trials 1 to 2 and from Trials 1 to 3. Although performance times decreased from Trials 2 to 3, this was not a significant decrease. In addition, a Subject Athletic Ability by Trials interaction f_(l,l58) (see Footnote 1) = 7.31, p_< .008, yielded significance. Post hoc analysis indicated that athletic and nonathletic subjects performed significantly longer on Trial 1 than on Trials 2 or 3. Also, athletic subjects extended their legs longer than nonathletic subjects on every trial. No other interactions with trials were significant. The results of this analysis for the experimental-only groups are summarized in Appendix C, Table 3. A separate ANOVA was conducted to compare the experimental subject modeling conditions to the athletic and nonathletic subject control groups. This was accomplished by means of a 2 x 2 x 3 (Experimental/ Control Groups x Subject Athletic Ability x Trials) ANOVA. Results of this analysis yielded a significant main effect for Subject Athletic Ability‘fi (1,146) = 38.94, p-< .0009. The means and standard deviations for the control groups are contained in Table 2. This main effect indi- cated that both experimental and control athletic subjects extended their legs significantly longer (M.= 100.42, §D_= 45.56) than nonathletic experimental and control subjects (M = 62.33,_§D = 27.36). This main 59 .mucmscoecmq co cowouaeapce apepwn< uwpm_;u< Homwnam an mow—Pa< UPSang< Paco: an Laucmu Paco: .4 «camel NEON-aflan 0. PNJIP.087) which showed that for female athletic subjects observing a female athletic model resulted in the longest performance; however, female athletic subjects observing a male athletic model exhibited the shortest performance. 62 Hypothesis 3. The third hypothesis stated that subjects viewing a model of similar athletic experience extended their legs longer than control subjects. To test this hypothesis for the athletic subjects, the averaged group motor performance score of the female athletic model and the male athletic model was compared to the averaged motor perform- ance score of the athletic control group. Results indicated no differ- ence between the two groups, t_(70) = .18. The third hypothesis for athletic subjects was, therefore, not supported. To test this hypothesis for the nonathletic subjects, the averaged group motor performance score of the female nonathletic model and the male nonathletic model was compared to the averaged motor performance score of the nonathletic control group. Results again indicated that the nonathletic subjects viewing a model of similar athletic experience did not differ significantly from nonathletic control subjects, 3 (70) = -2.30. Therefore, the third hypothesis for nonathletic subjects was not supported. Questionnaire Results Two questionnaires were administered to each subject in this study; a pre-experimental and a post-experimental questionnaire (see Appendices A and B). These questionnaires obtained background informa- tion regarding subjects' past competitive sports history as well as information concerning subjects' present involvement in any non-organized sports on a regular basis. In addition, subjects were asked whether or not they had any leg injury. Any subject indicating a leg injury was eliminated from the study. None of the Volunteer subjects indicated any 63 leg injuries. The questionnaires were also used to assess level and strength of subjects' self-efficacy, as well as subjects' opinions of whether males or females would perform best on the leg-extension task. Subjects exposed to the modeling conditions were asked to rate their perceived similarity to the model and to provide an explanation for their ratings. Subjects exposed to the modeling conditions were also asked to compare their performance to the model's performance, whether or not subjects felt that the model had an influence on their own per- formance (and why) and whether or not subjects competed with the model. The questionnaire results are divided into self-efficacy results; model/observer similarity results; descriptive analysis results concern- ing model/observer competition, perceived model influence, model/observer motor performance comparisons; and lastly, background information results. Self-efficacy results. On the efficacy questionnaires, Items 3, 4 and 5 were assessed for both level and strength of self-efficacy. The level of efficacy was assessed by asking the subject to rate her athletic ability a) in general, b) on muscular endurance tasks, and c) specifically on the leg-extension task, on 7-point, likert-type scales. Strength of efficacy was assessed by asking the subject to rate how certain she was of her athletic ability a) in general, b) on muscular endurance tasks, and c) on the experimental task, from zero to 100% certainty. 0n the pre-experimental questionnaire, self-efficacy Items 3, "general athletic ability" and 4, "ability to perform muscular endurance tasks" were each analyzed by means of a Model Gender x Model Athletic Ability x Subject Athletic Ability (2 x 2 x 2) ANOVA for both level and 64 strength of self-efficacy for the experimental-only groups. Results for the level of self-efficacy concerning subjects' general athletic ability and ability to perform muscular endurance tasks, indicated a significant main effect for Subject Athletic Ability on Item 3, f_(l,112) = 170.15, p_< .0009 and on Item 4, [_(l,112) = 84.78, p_< = .0009. These main effects indicated that athletic subjects possessed higher pre-performance levels of efficacy concerning their motor abili- ties than the pre-performance efficacy level of the nonathletic sub- jects. Specifically, athletic subjects were more confident in their general athletic ability (M = 6.00, SD = .78) and their ability to per- form muscular endurance tasks (M = 5.27, SD = .88) than were the non- athletic subjects (M = 3.58, S_D = 1.23; M = 3.53, E = 1.14, respec- tively). In addition, a significant main effect for Model Gender was evidenced on Item 3, general athletic ability, f_(l,112) = 5.06, p_< .026. Subjects who were exposed to a female model indicated a higher level of perceived general athletic ability (M_= 5.00, SD = 1.47) than the level of perceived general athletic ability indicated by subjects exposed to a male model (M = 4.58, E = 1.68). Results for strength of self-efficacy, again indicated a signifi- cant Subject Athletic Ability main effect for general athletic ability If (1,112) = 139.86, B.‘ .0009, and ability to perform muscular endurance tasks F (1,112) = 73.80, p_< .0009. In conjunction with the level of self-efficacy results, strength of self-efficacy results also revealed that athletic subjects were more certain of their general athletic ability (M'= 8.13, §Q_= 1.36) and their ability to perform muscular endurance tasks (M_= 7.07, SQ.= 1.68) than were the nonathletic subjects 65 for certainty of general athletic ability (M,= 4.35, SD = 2.11) and certainty of ability to perform muscular endurance tasks (M = 4.15, .SQ = 2.02). In addition, a significant main effect for Model Gender .5 (1,112) = 5.50, p_< .021 was again evidenced for strength of general athletic efficacy expectations. Subjects who were exposed to a female model, indicated a stronger degree of perceived ability to perform muscular endurance tasks (M = 6.62, SD = 2.50) than subjects exposed to a male model (M = 5.87, SD = 2.65). Results of these analyses for the experimental-only groups, are summarized in Appendix 0, Tables 5, 6, 7 and 8. A separate ANOVA was conducted to compare the experimental subject modeling conditions to the athletic and nonathletic subject control groups for efficacy Items 3 and 4. This was accomplished by means of a 2 x 2 (Experimental/Control Groups x Subject Athletic Ability) ANOVA for both level and strength of self-efficacy. Results for the level of self-efficacy concerning subjects' general athletic ability and ability to perform muscular endurance tasks also indicated a significant main effect for Subject Athletic Ability. On Item 3,.5 (1,146) = 187.97, _p < .0009 and on Item 4,.[ (1,146) = 86.22, p_< .0009. These main effects indicated that athletic subjects were more confident in their general athletic ability (M = 6.00, _SD = .84) and their ability to per- form muscular endurance tasks (M_= 5.20, §Q_= .93) than were the non- athletic subjects (M_= 3.67, SD = 1.21) (M_= 3.64, SD = 1.13) respec- tively. In addition, an Experimental/Control Groups by Subject Athletic Ability interaction F (1,146) = 4.26, p_< .041, yielded significance on Item 4. However, post hoc analyses (using Behren's Fisher 3' test for 66 unequal n's) indicated that viewing a model did not change level of efficacy for muscular endurance of the athletic subjects (Experimental Group M = 5.27, SD = .83; Control Group M = 4.93, SD = 1.10) or of the nonathletic subjects (Experimental Group M_= 3.53, SD = 1.04; Control Group M = 4.07, SD = 1.03). Experimental and control athletic subjects again had higher levels of efficacy for muscular endurance than experi- mental and control nonathletic subjects. Results for strength of self-efficacy in the experimental versus control groups ANOVA, indicated a significant Subject Athletic Ability main effect for general athletic ability [_(1,146) = 150.19, p_< .0009, and ability to perform muscular endurance tasks 5 (1,146) = 75.15, .p < .0009. In conjunction with the level of self-efficacy results, strength of self-efficacy results also revealed that athletic subjects were more certain of their general athletic ability (M.= 8.09, §Q_= 1.41) and their ability to perform muscular endurance tasks (M = 6.92, .SQ = 1.71) than were the nonathletic subjects for certainty of general athletic ability (M = 4.56, SD,= 2.09) and certainty of ability to perform muscular endurance tasks (M = 4.35, SQ.= 1.96). In addition, a significant interaction between Experimental/Control Groups and Subject Athletic Ability F (1,146) = 5.35, p_< .022 was again evidenced for strength of muscular endurance efficacy expectations. Post hoc analyses (using Behren's Fisher 3} tests) again indicated that for athletic subjects, viewing a model (M_= 7.07, SD = 1.14) did not increase the strength of their efficacy to perform muscular endurance tasks compared to controls (M_= 6.33, SD = 1.76). However, Viewing a model did decrease the strength of efficacy for nonathletic subjects 67 (M = 4.15, SD = 1.96) compared to controls (M = 5.13, _S_l_)_ = 1.55). Again, athletic experimental subjects had higher efficacy strength scores than nonathletic experimental subjects on muscular endurance measures. Results for general athletic ability efficacy strength re- vealed a trend (2 > .085) towards this same interaction. Results of these analyses for the experimental and control subjects are summarized in Appendix 0, Tables 9, 10, 11 and 12. Self-efficacy Item 5 which asked, "How well do you think your over- all performance will be on this task?" on the pre-experimental question- naire and "How well do you think your overall performance would be on this task in the future?" on the post-experimental questionnaire, was analyzed by means of a 2 x 2 x 2 x 2 (Model Gender x Model Athletic Ability x Subject Athletic Ability x Pre/Post Scores) ANOVA for both level and strength of self-efficacy for the experimental-only groups. Results for level of self-efficacy indicated a main effect for Subject Athletic Ability F(l,112) = 86.80, p_< .0009. This main effect indi- cated that athletic subjects possessed higher levels of self-efficacy concerning the experimental task across both pre-and post-experimental measures (M_= 5.20, SD = .83) than the level of self-efficacy (across pre- and post-measures) of nonathletic subjects (M = 3.70, g = .92). Strength of self-efficacy results for Item 5 also indicated a Subject Athletic Ability main effect 5 (1,112) = 48.02, P < .0009. Athletic subjects possessed stronger efficacy feelings (M = 6.77, SD = 1.67) across both pre- and post-strength of efficacy measures than the strength of self-efficacy (across both pre- and post-measures of the nonathletic subjects (fl.= 4.57, SD = 1.76). In addition to the Subject Athletic Ability main effect, a significant Pre/Post main effect 68 (p_< .009) was evidenced for strength of self-efficacy [_(1,112) (see Footnote 1) = 7.08. Specifically, strength of self-efficacy increased from pre-experimental measures of self-efficacy (M = 5.48, SD = 2.19) to post-experimental measures of self-efficacy (M_= 5.86, SD = 2.19). More interestingly, however, these main effects were superseded by a three-way (Model Gender x Subject Athletic Ability x Pre/Post Measures) interaction_§ (1,112) (see Footnote 1) = 5.90, p_< .017. Post hoc analysis (Tukey WSD) of this interaction indicated that there were no significant differences for athletic subjects; however, for nonathletic subjects. viewing a female model produced stronger efficacy expectations (fl = 4.87, _s_o (M = 3.83, so 1.94) prior to performance than viewing a male model 1.91). Also, nonathletic subjects who saw a male model, significantly increased their efficacy expectations after performing the task (M_= 4.87, §Q_= 2.24) (see Figure 5). The self-efficacy results of Item 5 for the experimental-only groups are summarized in Appendix D, Tables 13 and 14. A separate ANOVA was conducted to compare the experimental subject modeling conditions to the athletic and nonathletic subject control groups for self-efficacy Item 5. This was accomplished by means of a 2 x 2 x 2 (Experimental/Control Groups x Subject Athletic Ability x Pre/Post Scores) ANOVA for both level and strength of self-efficacy. Results for level of self-efficacy again indicated a main effect for Subject Athletic Ability E (1,146) = 77.23, p_< .0009. This main effect indicated that athletic subjects possessed higher levels of self-efficacy across both pre- and post-efficacy measures (M = 5.12,_§D = .91) than the level of self—efficacy (across pre- and post-measures) of nonathletic 69 .m EmpH co gpmcmcpm aumuwmem so» cowpuwcmch mmgammmz pmoa\mcm an hawpwn< umumpgp< pomwaam x: smegma Paco: whomaam OPPWJIP .063) toward the Subject Athletic Ability by Pre/Post Measures interaction. This interaction was mainly due to the self- efficacy scores of nonathletic subjects. Nonathletic subjects tended to exhibit a decrease in certainty of their ability to perform from 72 pre- (M = 43.35, §D_= 28.07) to post-experimental (M = 32.58, §Q_= 18.79) measures of efficacy strength. There was very little difference in strength of efficacy measures between the pre- (M_= 60.50, SD = 26.77) and post-measures (M = 58.17, SD = 26.71) obtained from the athletic subjects. For a summary of the 15—time increment efficacy strength results for the experimental-only group, see Appendix 0, Table 17. A separate ANOVA was conducted to compare the experimental subject modeling conditions to the athletic and nonathletic subject control groups on self-efficacy with regard to the 15 specified time lengths. The responses indicated by the experimental and control group subjects to these 15 questions, were summed to obtain a single score and submitted to an Experimental/Control Groups x Subject Athletic Ability x Pre/Post Measures (2 x 2 x 2) ANOVA. Results, again revealed a significant main effect for Subject Athletic Ability E = (1,146) = 24.26, p_< .0009. Athletic subjects possessed stronger self-efficacy feelings toward per- forming the task (M = 58.19, SD = 22.11) across both pre- and post- measures of self-efficacy strength than nonathletic subjects (M = 40.20, .SQ = 22.90). A significant Pre/Post main effect was evidenced f_(l,l46) (see Footnote 1) = 6.21, p.< .014 for strength of self-efficacy similar to the Pre/Post main effect found for the experimental-only groups. Specifically, self-efficacy strength decreased from pre- (M = 51.74, SD = 28.29) to post-experimental measures (M = 46.65, g = 26.23) across all subjects; however (as also noted among the experimental-only groups), this Pre/Post main effect is probably better interpreted by examining 73 the significant Subject Athletic Ability by Pre/Post Measures inter- action 5 = (1,146) (see Footnote 1) = 4.22, p_< .042. Post hoc analysis indicated that this analysis was again, mainly due to the self-efficacy scores of the nonathletic subjects. Nonathletic subjects exhibited a decrease in certainty of their ability to perform from pre- (M = 44.84, §Q_= 29.16) to post-experimental (M = 35.56, SD = 22.21) measures of efficacy strength. There was very little difference in strength of efficacy measures between the pre- (M 58.64, £1 = 25.79) and post- measures (M = 57.75, SD = 25.35) obtained from the athletic subjects. Again, athletic subjects had higher efficacy scores than nonathletic subjects on both pre- and post-measures. Adding the control subjects to this analysis provided the additional power to make the trend toward a significant interaction found for the experimental-only groups analy- sis, significant. For a summary of the 15 time-increment-efficacy strength results for experimental and control group subjects, see Appendix D, Table 18. Model/observer similaritygresults. Each subject exposed to a model was asked to rate her perceived similarity to that model on a 7-point, likert-type scale with 1 representing "very dissimilar" and 7 representing "very similar". In addition to this rating, subjects were provided with an open-ended item asking them to explain the reason(s) for their ratings. Responses to the ratings were analyzed by means of a 2 x 2 x 2 (Model Gender x Model Athletic Ability x Subject Athletic Ability) ANOVA for perceived model/observer similarity. Results from this analy- sis indicated that the main effect for Model Gender was significant 74 f_(l,112) = 6.36, p_< .013. This main effect indicated that female models were perceived to be more similar to the subjects (M = 3.60, .SQ = 1.61) than were the male models (M_= 3.00, SD = 1.50). A Subject Athletic Ability main effect was also significant £_(1,112) = 24.04, p_< .0009. Athletic subjects perceived themselves to be more similar to the models (M,= 3.88, SD = 1.51) than did the nonathletic subjects (M’= 2.72,_§Q = 1.43). The Subject Athletic Ability main effect is better interpreted by examining the significant interaction between Model Athletic Ability and Subject Athletic Ability with the frvalue being 26.86, df = 1,112, E.< .0009. Post hoc analysis using a Tukey WSD procedure (Winer, 1971) indicated that athletic subjects perceived themselves as being more similar to the athletic model (M.= 4.30, SD = 1.53) than to the non- athletic model (M = 3.47, §Q_= 1.38). In addition, nonathletic subjects indicated that they were more similar to the nonathletic model (M = 3.53, _S_D = 1.38) than to the athletic model (M = 1.90, SD = .92). The results of this analysis are summarized in Appendix D, Table 19. Descriptive statistics were used to help interpret the reason(s) given by the experimental subjects for their similarity ratings. Frequency distributions indicated that the most frequent response of all subjects exposed to the modeling conditions dealt with "physical items". Specifically, 69.17% of all respondents mentioned athletic experience, athletic background, present level of physical activity and/or physical appearance characteristics, excluding gender responses. None of the subjects indicated gender-only as a response, however, gender responses in combination with other items (physical and/or nonphysical, e.g., 75 time factors, student status, attitudes, etc.) accounted for 13.33% of all responses. An additional 10.83% of subject responses dealt with responses unrelated to gender or physical qualities. The remain- ing 6.67% of responses were not coded (e.g., no answer, etc.). For subjects viewing a male model, 63.33% mentioned athletic experience, athletic background, present level of physical activity and/or physical appearance characteristics alone or in combination with other "non-physical" items (excluding gender responses); a higher per- centage of subjects that observed a female model (81.67%) indicated some component of the "physical" qualities alone or in combination with other "non-physical" items (excluding gender responses). When considering subject athletic ability, descriptive statistics indicated that athletes (80%) tended to indicate at least one physical/ athletic item (excluding gender responses) more often than nonathletic subjects (73.83%). Only 17% of the nonathletic subjects indicated gender responses, while even fewer athletic subjects (11.67%) indicated gender as a reason for their similarity ratings. Somewhat related to the similarity/dissimilarity dimension was the post-experimental questionnaire item requiring experimental and control subjects to indicate who they thought would perform best on the leg-extension task, males or females. Descriptive statistics were again used to interpret subjects' responses. Frequency distributions indi- cated that the most frequent response was males which was chosen by 72% of all subjects. Only 15.33% of all subjects indicated that they thought females would perform best on the leg-extension task. Although subjects were only provided with a choice of male or female, 12.67% of all 76 subjects checked both male and female or wrote both for their responses to this question. Nonathletic subjects indicated that males would perform best on this task (73.33%) slightly more than did athletic subjects (70.67%). In addition, nonathletic subjects tended to have slightly higher expec- tations for females (18.67%) than did the athletic subjects (12%). However, substantially more athletic subjects (17.33%) indicated both as a response to this question than nonathletic subjects (8%). 0f the subjects exposed to a male model, 70% indicated that males would perform best on the leg-extension task. Interestingly enough, of the subjects exposed to a female model, 70% also indicated that males would perform best on this task. Subjects exposed to a female model held higher performance expectations for females (21.67%) than did subjects exposed to a male model; only 10% of the subjects exposed to a male model indicated that females would perform best on this task. Lastly, a higher percentage of subjects that viewed a male model indi- cated both as a response (20%) compared to the lower percentage of subjects indicating both who were exposed to a female model (8.33%). When examining experimental and control subjects separately, it was found that subjects who did not view a model (controls) thought that males would perform best on this task (80%) more often than did the subjects exposed to a model (70%). Control subjects also held lower expectations for females (13.33%) than did the experimental subjects (15.83%). Competition, influence and comparison results. All of the items in this section were analyzed by means of frequency distributions. 77 Only those subjects exposed to the modeling conditions were asked whether or not they competed with the person who demonstrated the leg- extension task. The majority of these subjects (62.5%) indicated that they did not compete with the model. Of the 37.5% of the subjects that did compete with the model, 46.67% of the athletic subjects com- peted, while only 28.33% of the nonathletic subjects competed. In examining the gender of the model that subjects were exposed to, athletic and nonathletic subjects did not drastically differ in the extent that they competed with male or female models, however, a slight difference did occur. Specifically, out of the athletic and nonathletic subjects that competed, slightly more competed with the female models (51.11%) than with the male models (48.89%). Both athletic and nonathletic subjects indicated that they competed more with the nonathletic models than with the athletic models. Out of the subjects that did compete, 50% competed with the nonathletic model, while only 25% of the subjects competed with the athletic model. Subjects exposed to the modeling conditions were asked whether or not they thought the model had an influence on their own performance. In addition, these subjects were provided with an open-ended item ask- ing them to explain the reason(s) for their responses. Descriptive statistics indicated that the majority of the experimental subjects (63.33%) felt that the model did not influence their own performances. 0f the 36.67% of the subjects that felt the model influenced their own performances, 40% of the athletic subjects, while 33.33% of the non- athletic subjects responsed in this manner. Of these same subjects, more thought that they were influenced by the nonathletic models (56.82%) 78 than the athletic models (43.18%) and more by female models (54.55%) than male models (45.45%). Descriptive statistics were also used to help interpret the reason(s) given by the subjects that indicated that the model did influence their own performances. Frequency distributions revealed that the most frequent response of the subjects answering "yes" to the influence question dealt with motivational factors. Fifty percent of these subjects felt that the model influenced their own performance by providing purely motivational cues. The next most frequent response provided by these subjects centered on informational factors. Of these responses, 29.55% included informational answers of two types: "information about how to perform the task", or "information about task difficulty". The third most frequent response offered by subjects (18.18%) was: "motivation and information about the model's condition and ability to perform the leg-extension task". The remaining 2.27% of the responses dealt with responses not coded as informational or motivational in nature (e.g., "basis for comparison"). In order to assess subjects' perceptions of their own performances in relation to that of the models', experimental subjects were asked, "How do you think the model performed in comparison to your performance?" Subjects were provided with a choice of one of three responses; "better", "same", or "worse". Descriptive statistics revealed that the most frequent response offered by subjects was that the model performed better (69.17%) in comparison to their own performances. Only 7.5% of the subjects felt that the model performed worse than themselves, while 23.33% felt that their performance was comparable to that of the model's. 79 Upon examining the responses in terms of subject athletic ability, it was found that 90% of the nonathletic subjects indicated that the model had performed better than themselves, while 48.33% of the athletic subjects responded in that manner. For the nonathletic subjects, the athletic models were thought to be slightly better than the nonathletic models compared to their own performances (male athletic model = 27.8%; female athletic model = 27.8%; male nonathletic model = 24%; female nonathletic model = 20.4%). However, for the athletic subjects, the male athletic model was more frequently thought to be a better performer (48.3%) than themselves, than were the other models (female athletic model = 27.6%; male nonathletic model = 13.8%; female nonathletic model = 10.3%). Background information results. All subjects were asked to com- plete pre-experimental background information questions. The first ques- tion required subjects to indicate their past (and present) organized, competitive sports experience. Subjects were provided with seven possi- ble response items: "youth sports team", "junior high school team", I'high school athletic team", "college athletic team", "outside league(s)", "college intramurals", or "none". Descriptive statistics revealed that 65.33% of the athletic subjects indicated involvement in youth sports and 16% of the nonathletic subjects indicated youth sports involvement. Junior high school athletic involvement was indicated by 76% of the athletic subjects and 14.67% of the nonathletic subjects. When con- sidering high school athletic experiences, none of the nonathletic sub- jects (0%) had particpated on high school teams, while 94.67% of the athletic subjects had experienced competitive high school athletics. 80 Outside athletic league involvement was indicated by 72% of the athletic subjects and none (0%) of the nonathletic subjects. Collegiate intra- mural participation was indicated by 14.66% of the nonathletic subjects and 41.33% of the athletic subjects. Lastly, the majority of nonathletic subjects (62.67%) indicated no form of athletic participation. The second background question asked subjects to indicate whether or not they were involved in any non-organized sport(s) on a regular basis. Descriptive statistics revealed that 58.67% of the athletic subjects and 24% of the nonathletic subjects regularly engaged in non- organized sports. The two most frequent forms of non-organized athletic participation indicated by athletic subjects were "jogging" (37.33%), followed by "lifting weights" (26.67%). The two most frequent forms of non-organized athletic participation indicated by nonathletic subjects were "jogging" (8%) and "swimming" (6.67%). Summary of ANOVA Results Motor performance ANOVAS indicated the following significant find- ings: 1. Athletic subjects extended their legs significantly longer than nonathletic subjects. 2. Subjects exhibited a significant decrease in performance times from Trials 1 to 2 and from Trials 1 to 3. 3. Athletic experimental and control subjects extended their legs significantly longer than nonathletic experimental and control subjects on the first trial only. Motor performance ANOVAS indicated the following trends: 1. Subjects observing a female athletic model tended to extend their legs longer than subjects observing a male athletic model. ings: 81 2. Athletic subjects observing a female athletic model tended to extend their legs longer than athletic subjects observing a male athletic model. Questionnaire ANOVAS indicated the following significant find- 1. Athletic subjects possessed higher pre-performance levels and strength of efficacy than nonathletic subjects, concern- ing a) general athletic ability, b) ability to perform muscular endurance tasks, c) overall ability to perform the leg extension task, and d) for the 15 specified time lengths. 2. Subjects who were exposed to a female model indicated higher levels and strength of pre-performance perceived general athletic ability than the pre-performance level and strength of perceived general athletic ability indicated by subjects exposed to a male model. 3. Subjects who were exposed to a female model indicated stronger pre-performance degrees of perceived ability to perform muscular endurance tasks than the strength of pre-performance perceived muscular endurance ability indicated by subjects exposed to a male model. 4. Strength of self-efficacy to perform the leg-extension task increased from pre- to post-experimental measures of self- efficacy; however, this main effect was superseded by a three-way interaction which indicated that there were no pre- post significant differences for athletic subjects; however, for nonathletic subjects, viewing a female model produced stronger efficacy expectations than viewing a male model. Also, nonathletic subjects who saw a male model, significantly increased their efficacy expectations after performing the task. 5. Self-efficacy strength decreased from pre- to post-experimental measures for the 15 specified time lengths; however, this was mainly due to the self-efficacy scores of nonathletic subjects. Nonathletic subjects exhibited a decrease in certainty of their ability to perform from pre— to post-experimental measures of efficacy strength. 6. Viewing a model did not change level or strength of efficacy for the performance of muscular endurance tasks of the athletic subjects or of the nonathletic subjects. 7. Viewing a model decreased the strength of efficacy to perform muscular endurance tasks compared to controls. 82 Model/observer similarity ANOVAS indicated the following signifi- cant findings: 1. Female models were perceived to be more similar to the sub- jects than were the male models. . Athletic subjects perceived themselves to be more similar to the models than did the nonathletic subjects. . Athletic subjects perceived themselves as being more similar to the athletic model than to the nonathletic model. . Nonathletic subjects indicated that they were more similar to the nonathletic model than to the athletic model. CHAPTER V DISCUSSION AND CONCLUSIONS The purpose of this study was to investigate the differential effects of model similarity across two dimensions, gender and perceived task ability, on the performance and self-efficacy of subjects on the leg-endurance task. A secondary aim was to investigate the effects of observers' athletic ability on modeling and self-efficacy of the leg-endurance task. It was hypothesized that a) female non- athletic subjects would perform the longest on the leg-endurance task after viewing the female nonathletic model, b) female athletic subjects would perform longest on the leg-endurance task after viewing either the female or male athletic model, and c) subjects who had observed a model of similar athletic ability would extend their legs longer than control subjects. No predictions were made concerning the most salient similarity cue, gender or perceived similarity in task ability, modeled by subjects due to a lack of previous knowledge concerning this issue. This chapter has been organized into three sections. The first section presents a discussion of this study and its findings. The second section states the conclusions of this investigation, while the third section offers implications and suggestions for future research. 83 —/LI [livi- II 84 Discussion This discussion has been organized in an effort to integrate the results of the previous modeling literature with the findings of this study. Questionnaire responses will be discussed in order to add clarity to specific performance findings. The primary purpose of this study which was to determine the most salient model/observer similarity cue, gender or perceived task ability, could not be determined for subjects' performance. Although subjects' perceptions of similarity were achieved across both dimensions, this perception of model/observer similarity did not influence their motor performance on the leg-extension task; this was especially apparent for nonathletic subjects. Perhaps one of the reasons for this inability to determine the most salient dimension of model/observer similarity was due to the lack of sufficient power necessary for detecting significant effects. The power of a test is defined as the probability of rejecting the null hypothesis. In order to test this plausible explanation, power for the Gould and Weiss (1981) investigation was calculated for the no-talk similar and dissimilar experimental groups. The resulting power, using Cohen's (1969) formula was .99. Power, in the present study, was calcu- lated separately for the athletic and nonathletic experimental groups. Power for the nonathletic treatment group, the group most comparable to the Gould and Weiss subject sample, was found to be .08. The power for the athletic treatment group was somewhat higher (.16). Possible explanations for the difference in power between the two studies could 85 be the Combination of the small sample size and large variability found in the present study. In terms of power, the larger the sample size, the smaller the sampling fluctuation of a statistic (e.g., standard error). By increas- ing the sample size, the standard error of the sampling distribution is reduced; in addition, the power to detect significant effects is increased. Gould and Weiss (1981) found significant motor performance differences between the nonathletic subjects exposed to the different modeling conditions. The lack of significant differences found among the nonathletic subjects in the present study may have been due to the relatively small number of nonathletic subjects assigned to each cell. In the experimental conditions, 15 nonathletic subjects were assigned to each one of the four modeling conditions or to the control condition. In the Gould and Weiss investigation, 60 nonathletic subjects were exposed to each one of the two modeling conditions with 30 nonathletic subjects assigned to the control condition. By extending the present investigation to include athletic as well as nonathletic subjects, and by extending the number of models from two to four, a condition was created in which the total sample size of the entire study may have needed to be substantially increased to realize experimental effects. However, the decision was made to test only 15 subjects per cell because that was the sample size per cell in the Gould and Weiss study. Regarding the variability factor, this investigation noted a great deal of variability for the experimental and control groups. This was clearly evident when examining the elevated Ms error terms for each of the analyses. Other investigations have also found high variability 86 when using the same leg-extension task (Corbin, Landers, Feltz & Senior, in press; Gould & Weiss, 1981; Martens & Landers, 1969; Weinberg, Gould & Jackson, 1979). A small sample size and large performance variability were probably also the reasons why the first and third hypotheses were not supported. The second hypothesis may not have been supported as specifically stated because the female athletic model group, which elicited the highest performance and the male athletic model group, which elicited the lowest performance were averaged together when performing the a priori contrasts, and this averaged score indicated that there were no significant differences between the two averaged groups. However, when the most similar group (female athletic model) was compared to all other groups in an a posteriori contrast there was still no significant dif- ference, t_(70) = 1.42, p_> .17. Although results of this study did not support any of the hypothe- ses as stated, a number of interesting performance trends did occur. For instance, the present investigation revealed a trend for athletic subjects which indicated that athletic subjects assigned to the female athletic model group tended to have longer performance times than sub- jects assigned to the male athletic group. A tentative explanation for this finding may be that female athletic subjects were less threatened by societal evaluations received when competing with a female model as compared to the possible negative social sanctions received if they had competed with a male model. In support of this notion is the question- naire data which revealed that more subjects felt influenced by the female models than the male models, and more subjects indicated that 87 they competed with the female models than with the male models. Another plausible explanation for this trend may have been the different motivation and encouragement provided by the models. Specifically, for female subjects, the female athletic model may have provided more motivation than the male athletic model, eliciting the attitude that "if she can do it, so can I". Indeed, subjects in the present study that reported being influenced by the models, seemed to provide support for the importance of motivation. The majority of responses as to why the model influenced subjects' performances indi- cated that the model provided motivational cues. 1 Perhaps the conclusion by Landers and Landers (1973) that modeling may, in fact, play a significant role in observer attention/motivation can be related to subjects' perceptions and the motor performance scores found in the present study. Landers and Landers (1973) found that ele- mentary school children who had observed an unskilled peer performed better than children who had observed an unskilled teacher. Possibly, subjects felt less threatened by the unskilled peer and hence, in more of a position to compete with the peer. It is possible that the chil- dren in the Landers and Landers study felt less threatened in terms of societal evaluation when a) competing against a fellow student rather than when competing against a teacher (authority figure) and b) when perceiving themselves to be in more of a position to emerge victorious when competing against an unskilled rather than skilled model. In addi- tion, a peer perceived as equal or less in ability to oneself, may introduce the "if he/she can do it, so can 1" philosophy. It must be 88 remembered that caution be observed when interpreting this finding because the differences were short of significance. Although no hypothesized differences for the effects of modeling on self-efficacy were made, some differential effects of model similarity were found. Even though the perception of model similarity existed along both model gender and model ability dimensions, similarity in terms of gender seemed to be more influential of self-efficacy. Female models elicited higher levels of efficacy and stronger efficacy perceptions of general athletic ability and the ability to perform muscular endurance tasks for athletic and nonathletic subjects than male models. Furthermore, for nonathletic subjects, female models also influenced overall present and future predictions for performance on the leg-extension task more than male models. However, there were no differences concerning model gender on the lS-item time lengths. No significant differences for model athletic ability were found when analyzing subject self-efficacy. The self-efficacy results of this investigation partially support the results of Gould and Weiss (1981) who found that similar model sub- jects (a female nonathletic model) had higher levels of self-efficacy and were more confident than dissimilar model subjects (a male athletic model), though they could not determine which similarity characteristic, gender or athletic ability, most greatly influenced subjects' efficacy appraisals. Again, the reason for Gould and Weiss' finding may be due to the subject's expectation that if another woman can do it, so can she. However, the present results must be taken in light of the fact that a comparison between the self-efficacy of subjects before they 89 viewed any one of the models and again after they had viewed the models was not made. Subjects' pre-performance efficacy was only measured after viewing the models and again after performing the task (post- efficacy). No efficacy measurement was taken prior to exposure to the models. This is limiting in the sense that it is not known to what degree observing a particular gender or particular athletic ability model changed subjects' initial degree of self-efficacy. A secondary aim of this investigation was designed to examine the effects of observers' athletic ability on modeling and self-efficacy of performing the leg-endurance task. For both motor performance and all self-efficacy measures, athletes had higher scores than nonathletes. Somewhat related to this finding is the fact that 90% of the nonathletic subjects indicated that the model performed better than themselves (perhaps reflecting a low degree of self-efficacy) while only 48.33% of the athletic subjects responded in that manner (perhaps reflecting higher perceptions of self-efficacy). One of the more interesting find- ings was the fact that in comparison to controls, viewing a model actually decreased nonathletic subjects' efficacy expectations for per- formance. In examining the individual group means of all subjects, the only nonathletic group that had higher efficacy scores than controls was the group exposed to a female nonathletic model. In addition, non- athletes also decreased their efficacy expectations after performing the task whereas athletes stayed about the same. Perhaps these findings can be related to the Weinberg et al., 1981 investigation. Weinberg et a1. (1981), postulated an interesting efficacy modeling effect. Specifically, Weinberg proposed that subjects 90 who were experimentally manipulated to achieve a high degree of self- efficacy responded to a specific failure situation by increasing their persistence efforts on the leg endurance task due to a state of internal cognitive dissonance created by their situational failure experience and their conflicting manipulated perceptions of high self-efficacy. Weinberg et a1. (1981), also suggested that subjects who were experimen- tally manipulated to achieve a low degree of self-efficacy responded to the specific failure situation by reinforcing their feelings of inade- quacy and convincing themselves that they would never perform adequately and hence, would fail in their future attempts when performing this task. Although the present study did not involve model or subject success or failure experimental evaluations, athletic subjects possessing a higher degree of self-efficacy than the nonathletic subjects, perhaps were not devastated by observing a model and did not change their efficacy expec- tations. However, possible self-defeating thoughts of the nonathletic subjects already possessing lower degrees of self-efficacy, may have dominated their cognitions to the extent that the observation of any successful model performing the experimental task reinforced their feel- ings of inadequacy. These subjects may also have convinced themselves that they would perform poorly in future attempts at this task, which may explain why they lowered their already depressed feelings of self- efficacy after performing the task. This postulation may explain the self-efficacy findings of the present study. Another plausible explanation for these results may have to do with the higher efficacy and perhaps higher competency perceptions noted among athletic subjects as proposed by Harter (1978). Harter (1978) proposes 91 that people who perceive themselves as competent (athletes in the present investigation) will persist at a task and possess a desire to master the task, while people who perceive themselves as incompetent (nonathletes in the present investigation) will exhibit lower persistence and a lower desire to master a task. Harter (1978) also noted that competency perceptions arise as a result of past, reinforcing successful experiences (i.e., athletes in the present study), which increases competency feelings, which in turn heightens the individuals' desire to persist at that task until it is mastered. A history of unsuccessful experiences or a history of little experience within a particular realm decreases competency feelings or does not provide the opportunity for the development of competency feelings,which in turn lowers the desire to master a task, lowering task persistence. The nonathletic subjects in this study had little experience with any form of athletics or physical activity. Perhaps, as Harter suggests, this is why their efficacy and performances were low. As a final note, athletes indicated that they competed more with the model than nonathletes. This finding supports Bandura's note (c.f., Gould & Weiss, 1981) that individuals who feel efficacious are likely to compete with others, whereas those who feel inefficacious avoid competi— tion. However, it was interesting, that of all the subjects who com- peted with a model, more competed with the nonathletic models than with the athletic models. This conflicts somewhat with Gould and Weiss' (1981) finding that similar model subjects competed more with the similar model than with the dissimilar model. They concluded that perceived similar- ity between model and observer may have heightened the social comparison 92 process, increasing observer motivation. From the present findings, however, this conclusion may not be tenable. It may be perceived superiority rather than perceived similarity on the part of the observer that increases observer motivation since athletic subjects in this study did compete more with the nonathletic than with the athletic models. Conclusions Based upon the findings and within the limitations of this study, the following conclusions were reached: 1. This investigation was unable to determine the more salient dimension of model similarity, gender or perceived task ability, influencing subjects' motor performance. 2. Gender is a more powerful similarity cue than perceived task ability for college-aged females' efficacy expectations on motor per- formance tasks. 3. Both gender and perceived task ability are influential simi- larity cues when considering subjects' perceptions of similarity. 4. Athletic subjects exhibited longer performances on the leg endurance task and higher degrees of self-efficacy than nonathletic subjects. c§uggestions for Future Research Several suggestions concerning future research on model similarity can be stated. First, all of the noted limitations of this study should be taken into account and hopefully rectified or controlled for in future studies. Expecially important is the future obtainment of two 93 pre-efficacy measurements, one before subjects view a model and a pre- measurement taken again after viewing the model but before task perform- ance. Only the latter pre-efficacy measurement was obtained in this study and the Gould and Weiss (1981) study. Due to the high variability that has been found with this task, future investigations employing this task may want to consider the use of pre-experimental performance as a covariate. Regarding sample size, the relatively small number of subjects per treatment cell probably masked any significant effects for the nonathletic as well as athletic subjects. If time and experimenter availability constraints did not exist, the total sample size of the entire study could have been substan- tially increased. Since lack of testing time and experimenter avail- ability was a problem, perhaps the use of athletic subjects should have been eliminated. The next logical step after the Gould and Weiss (1981) investiga- tion should have been the addition of two more models (athletic female, nonathletic male) without the addition of athletic subjects. After the study was extended to include four models, the next step could have involved the testing of athletic subjects to see if the findings generalized to that population. REFERENCES REFERENCE NOTES 1. Landers, D. M., & Landers, D. M. Modeling and motor behavior: A review of current theories and research. Unpublished manu- script,*l976. 94 REFERENCES Allpggz, F. H. Social psychology. Cambridge, Mass.: Riverside Press, Aronfreed, J. The problem of imitation. In L. P. Lipsitt 8 H. W. Reese (Eds.), Advances in child development and behavior, Vol. 4. New York: Academic Press, 1969. Bandura, A. Principles of behavior modification. New York: Holt, Rinehart & Winston, 1969. Bandura. A. Psychological modeling: Conflicting theories. New York: Aldine-Atherton,7197l. Bandura, A. Aggression: A social learning analysis. Englewood-Cliffs, N. J.: Prentice-Hall, 1973. Bandura, A. Self-efficacy: Toward a unifying theory of behavioral change. Psychological Review, 1977a, 84, 191-215. Bandura, A. Social learning theory. Englewood Cliffs, N. J.: Prentice-Hall, 1977b. Bandura, A. Reflections on self-efficacy. In S. Rachman (Ed.), Advances in behavior research and therapy Vol. I. Oxford: Pergamon Press, 1978. Bandura, A., & Adams, N. E. Analysis of self-efficacy theory of behavioral change. Cognitive Therapy and Research, 1977, 1, 287—310. Bandura, A., Adams, N. E., & Beyer, J. Cognitive processes mediating behavioral changes. Journal of Personality and Social Psychology, l977,_§§, 125-139. / Bandura, A., & Barab, P. G. Processes governing disinhibitory effects through symbolic modeling. Journal of Abnormal Psychology, 1973, 82, 1-9. Bandura, A., Blanchard, E. G., & Ritter, B. The relative efficacy of desensitization and modeling approaches for inducing behavioral, effective, and attitudinal changes. Journal of Personality and Social Psychology, 1969, 13, 173-199. 95 96 Bandura, A., & Huston, A. C. Identification as a process of incidental learning. Journal of Abnormal and Social Psychology, 1961, 63, 31 -3 8. _— Bandura, A., & Kupers, C. J. Transmission of patterns of self-reinforce- mgg: through modeling. Journal of Abnormal and Social Psychology. , 62, 1-9. Bandura, A., & Menlove, F. L. Factors determining vicarious extinction of avo1dance behavior through symbolic modeling. Journal of Person- gljty and Social Psychology, 1968, 6, 99-108. ‘T Bandura, A., Ross, 0., & Ross, S. A. Imitation of film-mediated aggres- sive models. Journal of Abnormal and Social Psychology, 1963, 66, 3-11. Bandura, A., & Walters, R. H. Adolescent aggression. New York: Ronald Press, 1959. Bandura, A., & Walters, R. H. Social learning and personality develop- ment. New York: Holt, Rinehart & Winston, 1963. Baron, R. A. Attraction toward the model and model's competence as determinants of adult imitative behavior. Journal of Personality and Social Psychology, 1970, 16, 345-351. Birrell, S. Achievement related motives and the women athlete. In C. A. Oglesby (Ed.), Women and sport: From myth to reality. Philadelphia: Lea & Febiger, 1978. Borkovec, T. D. The role of expectancy and physiological feedback in fear research: A review with special reference to subject character- istics. Behavior Therapy, 1973, 4, 491-505. Brown, I., Jr., & Inouye, D. K. Learned helplessness through modeling: The role of perceived similarity in competence. Journal of Person- ality and Social Psychology, 1978, 66, 900-908. Cohen, J. Statistical power analysis for the behavioral sciences. N.Y.: Academic Press, 1969. Corbin, C. 8., Landers, D. M., Feltz, D. L., 8 Senior, K. Sex differ- ences in performance estimates: Female lack of confidence vs. male boastfulness, Research Quarterly, in press. Deaux, K., & Ferris D. Attributing causes for one's.performance: The effects of sex, norms, and outcome. Journal of Research in Person- ality, 1977, 11, 59-72. Feldman-Summers, S. A., & Kiesler, S. 8. Those who are number two try harder: The effects of sex on attributions of causality. Journal of Personality and Social Psychology, 1974, 66, 846-855. 97 Feltz, D. L., Landers, D. M., & Raeder, U. Enhancing self-efficacy in high avoidance motor tasks: A comparison of modeling techniques. Journal of Sport Psychology, 1979, 1, ll2-122. Flanders, J. P. A review of research on imitative behavior. Psychological Bulletin, 1968, 69, 316-337. Gerst, M. D. Symbolic coding processes in observation learning. Journal of Personality and Social Psychology, l97l, lg, 7-l7. Gewirtz, J. L., & Stingle, K. G. Learning of generalized imitation as the basis for identification. Psychological Review, 1968, 16, 374-397. Gould, D., & Roberts, G. C. Modeling and motor skill acquisition. Quest, 1982, 66, 214-230. Gould, D., & Weiss, M. The effects of model similarity and model talk on self-efficacy and muscular endurance. Journal of Sport Psychology, 1981, 6, 17-29. Harter, S. Effectance motivation reconsidered: Toward a developmental model. Human Development, 1978, g1, 34-64. Hetherington, E. M., & Frankie, G. Effects of parental dominance, warmth, and conflict on imitation in children. Journal of Personality and Social Psychology, 1967, 6, 119-125. Hicks, D. J. Imitation and retention of film-mediated aggressive peer gnd adult mogels. Journal of Personality and Social Psychology, 965, g, 97- 00. House, W. C. Actual and perceived differences in male and female expec- tancies and minimal goal levels as a function of competition. Journal of Personality, 1974, fig, 493-509. Kazdin, A. E. Covert modeling and the reduction of avoidance behavior. Journal of Abnormal Psychology, 1973, 61, 87-95. Kazdin, A. E. Covert modeling, model similarity, and the reduction of avoidance behavior. Behavior Therapy, 1974, 6, 325-340. Kidd, T. R., & Woodman, W. F. Sex and orientations toward winning in sport. Research Quarterly, 1975, 66, 476-483. Klinger, E. Modeling effects on achievement imagery. Journal of Per- sonality and Social Psychology, 1967, 1, 49-62. Kohr, R. L. A comparison of statistical procedures for testing Ul=U2 with unequT’ns andivariances. Unpublished Ph.D. dissertation. The Pennsylvania State University (1970). 98 Kornhaber, R. C., & Schroeder, H. E. Importance of model similarity on extinction of avoidance behavior in children. Journal of Consulting and Clinical Psychology, 1975, 66, 601-607. Landers, D. M., & Landers, D. M. Teacher versus peer models: Effects of model's presence and performance level on motor behavior. ggurnal of Motor Behavior, 1973, 6, 129-139. Lenney, E. Womens' self-confidence in achievement settings. Psycholggjcal Bulletin, 1977,.§fl, 1-13. Lenney, E., Browning, C., & Mitchell, L. What you don't know can hurt you: The effects of performance criteria ambiguity on sex differences in self-confidence. Journal of Personality, 1980, 66, 306-322. Lewis, S. A comparison of behavior therapy techniques in the reduction of fearful avoidance behavior. Behavior Therapy, 1974, 6, 648-655. Martens, R., Burwitz, L., & Zuckerman, J. Modeling effects on motor performance. The Research Quarterly, 1976, 67, 277-291. Martens, R., & Landers, D. M. Coaction effects on a muscular endurance task. Research Quarterly, 1969, 66, 733-737. May, J. G. A developmental study of imitation. Dissertation Abstracts, 1966, 26, 6852-6853. McHugh, M. C., Duquin, M. E., & Frieze, I. H. Beliefs about success and failure: Attribution and the female athlete. In C. A. Oglesby (Ed.), Women and Sport: From myth to reality. Philadelphia: Lea & Febiger, 1978. Meichenbaum, D. H. Examination of model characteristics in reducing avoidance behavior. Journal of Personality and Social Psychology. 1971,_17, 298-307. O'Connell, E. J., Jr. The effect of cooperative and competitive set on the learning of imitation and nonimitation. Journal of Experi- mental and Social Psychology, 1965, 1, 172-183. Ogilvie, 8. C. The personality of those women who have dared to succeed in sport. In J. H. Goldstein (Ed.), Games, play & sport. Lea & Febiger, 1978. Perry, 0. G., & Perry, L. C. Observational learning in children: Effects of sex of model and subjects sex role behavior. Journal of Personality and Social ngchology, 1974, 61, 1083-1088. Reichard, G. A. Social life. In F. Boas (Ed.), General Anthropology. Boston: Health, 1938. 99 Roberts, J. M., & Sutton-Smith, 8. Child training and game involvement. Ethnology, 1962, 1, 166. Rosekrans, M. Imitation in children as a function of perceived simi- larity to a social model and vicarious reinforcement. Journal of Personality and Social Psychology, 1967, 1, 307-315. Rosenbaum, M. E., & Tucker, I. F. The competence of the model and the learning of imitation and nonimitation. Journal of Experimental Psychology, 1962, 66, 181-190. Weinberg, R. S., Gould D., & Jackson A. Expectations and performance: An empirical test of Bandura's self-efficacy theory. Journal of Sport Psychology, 1979, 1, 320-331. Weinberg, R. S., Gould, D., Yukelson, D., & Jackson, A. The effect of preexisting and manipulated self-efficacy on a competitive muscular endurance task. Journal of §port Psychology, 1981, 6, 345-354. Weinberg, R. 5., Sinardi, M. & Jackson, A. Effect of bar height and modeling on anxiety, self-confidence and gymnastic performance. International Gymnast, 1982, 6, TS, 11-13. Weinberg, R. S., Yukelson, D., & Jackson, A. Effect of public and private efficacy expectations on competitive performance. Journal of Sport Psychology, 1980, 6, 340-349. Winer, B. J. Statistical principles in experimental design. New York: McGraw-Hill,11971. Zimmerman, B. J., & Blotner, R. Effects of model persistence and success on children's problem solving. Journal of Educational Psychology, 1979, 11, 508-513. FOOTNOTES 1The lambda hat adjustment for degrees of freedom was used to correct for non-symmetry in the variance-covariance matrix. 100 APPENDICES APPENDIX A PRE-EXPERIMENTAL QUESTIONNAIRE APPENDIX A PRE-EXPERIMENTAL QUESTIONNAIRE Subject No. Group la. What 15 your past competitive sports experience? (check all that apply Youth Sports Team Junior High School Team _______ High School Athletic Team ______ College Athletic Team Outside League(s) College Intramurals None b. 00 you engage in any non-organized sports on a regular basis? (i.e., jogging, swimming, weightlifting, etc.) yes; if yes please explain 1'10 2. Do you have any leg injury that could affect your performance on this task? yes no 101 Rate On The Scale Below: 102 % of certainty of your ability to perform: very uncertain somewhat certain very certain O 10 20 30 4O 50‘ 601 70 80 90 100 your general athletic ability. 3-%* f %- / %’ / l 1 2 3 4 5 6 7 poor average good your ability to perform uscular endurance tasks. 4.; x 7 / T' x x l 2 3 4 5 6 7 poor average good how well you think your overall performance will be on this task. I I I I l fir 5] 7 l I II I l 2 3 4 5 6 7 poor average good Place a (Y) on the appro- priate line if you will be able to extend your leg aboVe the cord for: % 0f Certainty Of Your Ability To Perform The Task 101 20 30 4O 50 60 7O 80 90 very uncertain somewhat certain very certain 0 100 30 sec.? 45 sec.? 60 sec.? «>0on . 1 min. 15 '0 SEC . 10. l min. 30 sec.? 11. l min. 45 sec.? 12. min.? 13. min. 15 sec.? 14. min. 30 sec.? 15. min. 45 sec.? 16. min.? wwmmmm 17. min. 15 sec.? 18. min. 30 sec.? on 19. 3 min. 45 sec.? 20 4 min.? APPENDIX B POST-EXPERIMENTAL QUESTIONNAIRE APPENDIX B POST-EXPERIMENTAL QUESTIONNAIRE Subject No. Group % Of Certainty Of Your Future Ability To Perform very uncertain somewhat certain very certain Rate On The Scale Below: 0 10 '20 130 40 50 60 70 80 90 100 how well you think your overall performance would be on this task in the future I l l l l l I r T I r I l l 2 3 4 5 6 7 poor average good Place a (v0 on the appro- priate line if you think that in the future you % 0f Certainty Of Your Ability To would be able to extend Perform your leg above the cord for: very uncertain somewhat certain very certain 0 1O 20 30 4O 50 60 7O 8O 90 100 30 sec.? sec. sec. m m 3 n. n. n. n. n. n. 11. 1'1. n. 11. n. Do you think males or females would perform best on this task? males females 103 104 How similar do you perceive yourself to be to the person who demon- strated this task? j J l l 1 l 1 I I T I 7 I F l 2 3 4 5 6 7 very somewhat very dissimilar similar similar WHY? How do you think the model performed in comparison to your performance? better same worse Do you think the model had an influence on your performance? yes no If yes, why? Did you compete with the person who demonstrated this task? That is, did you try to do better than they did? yes no APPENDIX C MOTOR PERFORMANCE DATA 105 Table 3 Analysis of Variance for Motor Performance Experimental-Only Subjects Source 6:_ .66 .5 Model Gender (A) 1 1195.378 .30 Model Athletic Ability (8) 1 243.378 .06 A x B 1 14668.900 3.651 Subject Athletic Ability (C) l 160444.400 39.87* A x C 1 816.011 .20 B x C 1 2064.011 .51 A x B x C 1 11971.600 2.98 ++ Error between 112 4024.179 Trials (J) 1 31332.790 49.78* A x J 1 1350.903 2.15 B x J 1 546.103 .87 A x B x J 1 221.908 .35 C x J 1 4602.869 7.31* A x C X J 1 50.119 .08 B x C x J 1 42.086 .07 A x 8 x C x J 1 50.358 .08 Error within 158 629.424 Satterthwaite Ms error 184 1761.009 * B.‘ .05 II.6 s .059 +lg > .087 106 Table 4 Analysis of Variance for Motor Performance of Experimental versus Control Subjects Source 6i: g _E Experimental/Control Groups (A) 1 2762.722 .66 Subject Athletic Ability (B) 1 163210.900 38.94* A x B 1 7688.000 1.83 Error between 146 4190.928 Trials (J) 2 38015.240 60.83* A x J 2 349.677 .56 B x J 2 4644.229 7.43* A x B x J 2 400.552 .64 Error within 222 624.929 Satterthwaite Ms error 235 1813.595 * B.‘ .05 APPENDIX D QUESTIONNAIRE DATA 107 Table 5 Analysis of Variance for Questionnaire Item 3a__ Level of Efficacy for Experimental-Only Subjects Source 6:_ '66 .5 Model Gender (A) 1 5.208 5.06* Model Athletic Ability (B) 1 1.008 .98 A x B 1 .208 .20 Subject Athletic Ability (C) 1 175.208 170.15* A x C 1 2.408 2.34 B x C l .408 .40 A x B x C 1 .008 .01 Error between 112 1.030 a3--"Rate your general athletic ability." {6 < .05 108 Table 6 Analysis of Variance for Questionnaire Item 3a_ Strength of Efficacy for Experimental-Only Subjects Source 61_ .66 .5 Model Gender (A) 1 16.875 5.50* Model Athletic Ability (B) 1 1.875 .61 A x 8 1 1.875 .61 Subject Athletic Ability (C) 1 429.408 139.86* A x C 1 7.008 2.28 B x C l .675 .22 A x B x C l .408 .133 Error between 112 3.070 a3--"Indicate your percent certainty of your general athletic ability." #6 < .05 109 Table 7 Analysis of Variance for Questionnaire Item 4a_ Level of Efficacy for Experimental-Only Subjects Source 6:. .66 .6 Model Gender (A) l .533 .50 Model Athletic Ability (B) 1 .033 .03 A x B l .833 .78 Subject Athletic Ability (C) 1 90.133 84.78* A x C 1 2.133 2.01 B x C l .033 .03 A x B x C l .033 .03 Error between 112 1.063 a4--"Rate your ability to perform muscular endurance tasks." f6_< .05 110 Table 8 Analysis of Variance for Questionnaire Item 4a- Strength of Efficacy for Experimental-Only Subjects Source 6:_ .66 .5 Model Gender (A) 1 7.008 2.03 Model Athletic Ability (B) l .075 .02 A x B 1 3.675 1.06 Subject Athletic Ability (C) 1 255.208 73.80* A x C 1 5.208 1.51 B x C 1 1.875 .54 A x B x C 1 .208 .06 Error between 112 3.458 a4--"Indicate your percent certainty of your ability to perform muscu- lar endurance tasks." f6_< .05 111 Table 9 Analysis of Variance for Questionnaire Item 3a_ Level of Efficacy for Experimental versus Control Subjects Source 6:_ .66 .6 Experimental/Control Groups (A) 1 1.042 .96 Subject Athletic Ability (B) 1 204.167 187.97* A x B 1 1.042 .96 Error between 146 1.086 a3--"Rate your general athletic ability." f6,< .05 112 Table 10 Analysis of Variance for Questionnaire Item 3a_ Strength of Efficacy for Experimental versus Control Subjects Source 61: pg 6 Experimental/Control Groups (A) 1 4.335 1.39 Subject Athletic Ability (8) 1 468.167 150.19* A x 8 1 9.375 3.01+ Error between 146 3.117 a3--"Indicate your percent certainty of your general athletic ability." fgi< .05 *2 > .085 113 Table 11 Analysis of Variance for Questionnaire Item 4a_ Level of Efficacy for Experimental versus Control Subjects Source fl _M_S_ _I: Experimental/Control Groups (A) 1 .240 .23 Subject Athletic Ability (B) 1 91.260 86.22* A x B 1 4.507 4.26* Error between 146 1.058 a4--"Rate your ability to perform muscular endurance tasks." {6 < .05 114 Table 12 Analysis of Variance for Questionnaire Item 4a_ Strength of Efficacy for Experimental versus Control Subjects Source .66 .66 f. Experimental/Control Groups (A) l .375 .11 Subject Athletic Ability (B) 1 248.327 75.15* A x B 1 17.682 5.35* Error between 146 3.304 a4--“Indicate your percent certainty of your ability to perform muscular endurance tasks." *6 < .05 Analysis of Variance for Questionnaire Item 5a_ Level of Efficacy for Experimental-Only Subjects 115 Table 13 Source 61: 65 6 Model Gender (A) 1 4.267 2.74 Model Athletic Ability (B) 1 1.067 .69 A x B 1 .267 .17 Subject Athletic Ability (C) 1 135.000 86.80* A x C l .267 .17 B x C 1 1.067 .69 A x 8 x C l .267 .17 Error between 112 1.555 Pre/Post Efficacy (J) l .017 .04 A x J l .150 .32 B x J 1 .017 .04 A x 8 x J 1 .417 .89 C x J l .417 .89 A x C x J l .150 .32 B x C x J 1 .417 .89 A x B x C x J 1 .817 1.74 Error within 112 .470 a5--Pre-efficacy: "Rate how well you think your overall performance will be on this task." Post-efficacy: "Rate how well you think your overall performance would be on this task in the future." {6 < .05 116 Table 14 Analysis of Variance for Questionnaire Item 5a_ Strength of Efficacy for Experimental-Only Subjects Source g 66 6 Model Gender (A) 1 13.067 2.16 Model Athletic Ability (B) 1 1.067 .18 A x B 1 1.667 .28 Subject Athletic Ability (C) 1 290.400 48.02* A x C l .067 .01 B x C 1 1.667 .28 A x 8 x C 1 .067 .01 Error between 112 6.048 Pre/Post Efficacy (J) 1 8.817 7.08* A x J 1 3.750 3.01 B x J 1 2.817 2.26 A x 8 x J 1 2.017 1.62 C x J l .150 .12 A x C x J 1 7.350 5.90* B x C x J 1 .817 .66 A x B x C x J l .817 .66 Error within 112 1.245 Satterthwaite 156 3.646 a5"P’e'9fficacy= "Indicate your precent certainty of how well you think your overall performance will be on this task." Post-efficacy: "Indicate your percent certainty of how well you think your overall performance would be on this task in the future." 52 < .05 117 Table 15 Analysis of Variance for Questionnaire Item 5a_ Level of Efficacy for Experimental versus Control Subjects Source 66. .66 ‘E Experimental/Control Groups (A) 1 .333 .20 Subject Athletic Ability (8) 1 128.053 77.23* A x 8 1 11.213 6.76* Error between 146 1.658 Pre/Post Efficacy (J) 1 .000 .00 A x J l .083 .16 B x J l .480 .94 A x 8 x J 1 .003 .01 Error within 146 .510 a5--Pre-efficacy: Post-efficacy: TE.‘ .05 "Rate how well you think your overall performance will be on this task." "Rate how well you think your overall performance would be on this task in the future." 118 Table 16 Analysis of Variance for Questionnaire Item 5a_ Strength of Efficacy for Experimental versus Control Subjects Source g g 6 Experimental/Control Groups (A) 1 .213 .04 Subject Athletic Ability (8) 1 292.053 51.40* A x B 1 15.413 2.71 Error between 146 5.682 Pre/Post Efficacy (J) 1 9.720 7.25* A x1] 1 .163 .12 B x J 1 .013 .01 A x B x J 1 .403 .30 Error within 146 1.340 a5--Pre-efficacy: "Indicate your percent certainty of how well you thiak your overall performance will be on this tas ." Post-efficacy: "Indicate your percent certainty of how well you think your overall performance would be on this task in the future." #6_< .05 Analysis of Variance for Questionnaire Item Assessing 119. Table 17 Strength of Efficacy for 15 Time Lengthsa of Experimental-Only Subjects Source .66 ‘66 .5 Model Gender (A) 1 2065.067 2.07 Model Athletic Ability (B) 1 522.150 .53 A x 8 1 8.817 .01 Subject Athletic Ability (C) 1 27392.070 27.52* A x C 1 123.267 .12 B x C 1 40.017 .04 A x B x C 1 2076.817 2.09 Error between 112 995.459 Pre/Post Efficacy (J) 1 2574.150 8.53* A x J 1 390.150 1.29 B x J 1 693.600 2.30 A x B x J 1 141.067 .47 c x .1 1 1066.817 3.541? A x C x J 1 104.017 .35 B x C x J 1 60.000 .20 A x B x C x J 1 201.667 .67 Error within 112 301.674 Satterthwaite 648.567 174 aStrength of Efficacy for 15 Time Lengths-- "Indicate your percent certainty of your ability to Pre-efficacy: extend your leg, prior to performing the task for the 15 time lengths." Post-efficacy: "Indicate your percent certainty of your ability to extend your leg again, in the future, on this task for the 15 time lengths." *2 < .05; 13> .063 120 Table 18 Analysis of Variance for Questionnaire Item Assessing Strength of Efficacy for 15 Time Lengthsa of Experimental versus Control Subjects Source 61: _M_S_ E Experimental/Control Groups (A) 1 358.613 .36 Subject Athletic Ability (B) 1 24282.000 24.26* AxB 1 “men an1 Error between 146 1001.044 Pre/Post Efficacy (J) 1 1940.563 6.21* A x J 1 642.403 2.06 B x J 1 1318.803 4.22* A x B x J 1 .163 .00 Error within 146 312.387 Satterthwaite 229 656.716 aStrength of Efficacy for 15 Time Lengths-- Pre-efficacy: "Indicate your percent certainty of your ability to extend your leg, prior to performing the task for the 15 time engths." ' Post-efficacy: "Indicate your percent certainty of your ability to extend your leg again, in the future, on this task." *2 < .05; 16> .067 121 Table 19 Analysis of Variance for Questionnaire Item Assessing Model/Observer Perceived Similarity Ratinga for Experimental-Only Subjects Source 66 _lfi 1: Model Gender (A) 1 10.800 6.36* Model Athletic Ability (8) 1 4.800 2.83 A x B l .133 .08 Subject Athletic Ability (C) 1 40.833 24.04* A x C 1 .033 .02 B x C 1 45.633 26.86* A x 8 x C 1 2.700 1.59 Error between 112 1.699 aModel/Observer Perceived Similarity Rating-- "Rate how similar you perceive yourself to be to the person who demonstrated this task on the 7-point scale provided." #6 < .05 APPENDIX E RAW DATA Card 1 Columns 1-3 4 5 9-10 11 12-13 APPENDIX E RAW DATA 1666 Subject number Blank Model gender Model athletic experience Subject athletic experience Blank Subject age Blank Subject major 122 Value Labels 1 = male, 2 = female, 3 = no model nonathletic l = athletic, 2 3 = no model 1 = athletic, 2 = nonathletic 01 = English 02 = Psychology 03 = Communications 04 = Physical Education/Recrea- tion/Coaching 05 = Health 06 = Math 07 = Zoology 08 = Engineering 09 = Business/Marketing/Accounting 10 = Criminal Justice 11 = No preference 12 = Computer Science 13 = Pre-Med./Vet., Medical Technology 14 = Hotel and Restaurant Manage- ment 15 = Nursing 16 = Social Work 17 = Nutrition/Dietetics Columns 12-13 cont'd 14 15 16 17 18 19-20 21-23 24 25-27 28 29-31 123 Blank Subject year in school Item Value Labels 18 = Counseling 19 = Speech and Audiology 20 = Advertising/Graphics 21 = Political Science 22 = Geology 23 = Social Science 24 = Special Education/Elementary Education 25 = Biology/Biochemistry/Micro- biology 26 = Packaging/Materials and Logistics 27 = Travel Agent/Tourism 28 = Theatre . 29 = Animal Science 30 = Pre-Law/Business Law 31 = Labor Relations/Public Affairs Management 32 = Geography 33 = Physiology 1 = freshman, 2 = sophomore, 3 = junior, 4 = senior, 5 = graduate student, 6 = other Blank Subject verbalization Subject nonverbal behavior Blank Trial 1; time in seconds Blank Trial 2; time in seconds Blank Trial 3; time in seconds 1 = yes, 2 l = yes, 2 no no Columns 32 33-38 39-40 41 42-43 44 45 46-47 48 49 50-51 52 53-54 55 124 Item Value Labels Blank Average time for three trials; time in seconds Blank l to 7; with 1 representing poor and 7 representing good Rate your general athletic ability Indicate your percent 00 to 10; with 00 representing certainty of your very uncertain and 10 represent- ability to perform ing very certain general athletic tasks Blank Rate your ability to l to 7; with 1 representing poor perform muscular en- and 7 representing good durance tasks Indicate your percent 00 to 10; with 00 representing certainty of your very uncertain and 10 represent- ability to perform ing very certain muscular endurance tasks Blank Rate how well you think your overall performance will be on this task 1 to 7; with 1 representing poor and 7 representing good Indicate your percent 00 to 10; with 00 representing certainty of your very uncertain and 10 represent- overall ability to ing very certain perform this task Blank 01 to 15; total number of pre- experimental, time increment check marks Level of efficacy Blank Columns 56-58 59 60 61-62 63 64-65 66 67-69 70 71-72 73-74 75-76 125 Item Strength of efficacy Blank Rate how well you think your overall performance would be on this task in the future Indicate your percent certainty of your overall ability to perform this task in the future Blank Level of efficacy Blank Strength of efficacy Blank Percent certainty of your ability influ- encing your perform- ance Percent certainty of luck influencing your performance Percent certainty of task difficulty influencing your performance Value Labels 000 to 150; sum of the first digits (first two digits for 100% column) of pre-experimental percent certainty check marks 1 to 7; with 1 representing poor and 7 representing good 00 to 10; with 00 representing very uncertain and 10 represent- ing very certain 01 to 15; total number of post- experimental, time increment check marks 000 to 150; sum of the first digits (first two digits for 100% column) of post-experimental percent certainty check marks 00 to 10; with 00 representing very uncertain and 10 represent- ing very certain 00 to 10; with 00 representing very uncertain and 10 represent- ing very certain 00 to 10; with 00 representing very uncertain and 10 represent- ing very certain Columns 77-78 79 80 Card 2 1-3 10 11 12 13 14 15 16 126 Item Value Labels Percent certainty of 00 to 10; with 00 representing your effort influ- very uncertain and 10 represent- encing your perform- very certain ance Blank Do you think males or 1 = males; 2 = females; (3 = females would per- both) form best on this task Subject number Blank Model gender 1 = male, 2 = female, 3 = no model Model athletic l = athletic, 2 = nonathletic experience 3 = no model Subject athletic 1 = athletic, 2 = nonathletic experience Blank What is your past competitive sports experience: Youth sports team 1 = yes, 2 = no Junior high school 1 = yes, 2 = no team High school athletic 1 = yes, 2 = no team College athletic team 1 = yes, 2 = no Outside league(s) 1 = yes, 2 = no College intramurals 1 = yes, 2 = no None 1 = yes, 2 = no Blank Columns 17 18 19 20 21 22 23-24 127 Item 00 you engage in any non-organized sports on a regular basis Blank Do you have any leg injury that could affect your perform- ance on this task Blank Rate how similar you perceive yourself to be to the person who demonstrated this task Blank 66y did you select the similarit rat- ing you inaicated Value Labels l = yes, 2 = no 1 = yes, 2 = no 1 to 7; with 1 representing very dissimilar and 7 representing very similar gender athletic background, experi- ence and physical activity level physical appearance leg shake student seriousness of performance age athletics/physical appearance/ student/age/seriousness of performance gender/athletic experience gender/physical appearance physical appearance/athletic experience task task/athletic experience gender/athletic experience/ physical appearance confidence, comfortable, relaxed attitudes lactic acid/athletic experi- ence time leg shake/athletic experience task/leg shake time/leg shake Columns 23-24 cont'd 25 26 27 28 29 30 31 32 128'. Item Blank How do you think the model performed in comparison to your performance Blank Do you think the model had an influ- ence on your performance Blank W_hy did the model influence your per- formance (if you selected "yes" for influence question) Blank Did you compete with the person who demon- strated this task? That is, did you try to do better than they did Value Labels 21 = gender/task/physical activity level 22 = everyone is different in abilities 23 = I don't know 24 = blank _1 ll better, 2 = same, 3 = worse 1 = yes, 2 = no (Blank if selected "no" for influence question) _a ll information about how to perform task motivation information about models condi- tion and ability information about task diffi- culty motivation and information about task difficulty = motivation and information about models condition and ability motivation and information about how to perform task 8 = basis for comparison 01 01 h (JON II \1 ll 1 = yes, 2 = no Columns 33 34 .1120 Blank team 129 Value Labels (blank if no team) volleyball swimming/diving gymnastics basketball softball tennis rugby \IO‘U'Ith-d II II II II II II II — N ~— - n- and—”-54..” N diode—floNo—m—HO-floN_MJfNO-‘ONNWNm—NNOWWNW—NMNWNd—‘NwdQ—N—O—O-‘w—Q—‘WW-°-‘-'-'-' “ N a _ u _ 2‘ N --N CorfffffNNddOmWWNN——O° Nr .‘_..__._._._._._.~N__._. NON—N-N—NONONdNON—NdN—N—N—————dO—O—O—O—OdOdO—O—O—O—O~O—O u N- F! N N N M N N ~— N N N N N N N N N N N N N ”A N N N N N P—O—O-‘Nfl—M-‘P-flo-‘-'-'—W:’MN“MNMNMNNNWNMNNN—fl—O—MWit—WN-NHMNr-AMNN—— NNN—NNNNNdN-‘NN-dNN—‘N ‘ § N .0—1 ~— —0 N ~ _n _6 Ann _6 _o N N INN N _n 1 N .l-I —..o -'—-—OOOOOOOOOOOOOOOOOO-v-t——————OOOOOOOOOOOOOOOOOOOSOO-—-—--'-——-'-'—OOOOOOOOOOOOOOOOOO WWmkWNWN—n—NMWOWOM—m-mr°WmP—WW OOWWMNN‘~&WNH~OMW§W~¢-O 130 132 067 076‘086333 608 507 608 10 058 608 16 106 10000910 2 109 102 112666 609 608 708 10 071 608 08 065 10010810 86 013loi72112666 609 710 609 10 100 508 10 090 09001009 08500 020 2 oh2333 507 608 507 11 082 607 05 062 10000110 393L06810§31051666 506 506 505 11 072 605 06 032 05020507 83‘061 0 1 0 3666 507 608 609 08 072 608 05 062 06010606 07 0 6 063 0 5666 709 606 605 05 033 605 06 027 07000007 00 108 068 0 2000 608 507 507 06 030 605 06 023 05050307 12 0 5‘0 3.0 0000 710 709 610 07 069 709 06 055 10001010 §832°102°§920§2333 608 507 605 07 039 605 07 051 03060708 160315321282120333 708 607 607 11 068 608 08 062 10000610 662016205220 6666 708 506 603 06 065 603 06 068 05000802 3866030106510 9333 506 506 605 11 052 605 11 058 05000500 38330172057 oglooo 608 605 606 06 015 608 10 069 05000807 11 086 072 o 0333 507 506 607 06 069 708 06 036 08000310 1 o 5 079 0 3000 101 101 100 03 016 200 05 025 01000505 166 o 7 066 098333 506 606 607 05 028 608 07 059 08030908 62 0102076 061000 605 606 606 06 052 606 06 067 08000000 1582°I°1°§° 036333 203 203 202 06 028 302 03 022 07000608 06130962122 085666 607 507 507 15 112 507 11 086 00000010 07220562063 058333 606 606 606 05 025 505 03 020 10000810 g3 20552030 061000 100 100 100 09 068 201 06 036 10000005 1123053‘062 077000 506 605 506 09 066 507 12 088 08050808 298 0631029 073000 606 606 606 06 028 606 03 025 10001010 100 029 0 0 069666 301 606 606 07 066 606 06 032 08000808 062 063 053 067333 201 302 202 03 008 303 03 021 07010507 056 0 2 036 050000 302 606 606 05 031 308 06 026 09000707 081 o 9 067 062333 306 306 306 06 010 306 06 007 06010506 06 019 021 029000 206 603 603 06 026 205 01 003 07000606 066 052 016 037333 606 606 606 06 030 605 06 020 05000005 060207610602067333 609 609 507 08 058 608 09 063 08000007 I31“126 103 119333 608 507 507 07 050 506 06 052 05000005 39§50‘D:0 61068000 608 608 506 O7 051 606 01 058 08050908 21509620 62120333 608 608 506 09 063 606 05 036 00000508 67 0 8 068 o 9666 709 508 508 07 055 607 06 021 08010708 27150 8 0 3 0 3666 608 506 506 16 073 506 16 089 06000707 g36l31223162320000 709 709 709 15 127 710 15 150 10000108 2 29 2136115112 6333 609 509 606 07 057 605 07 060 09000909 2 I08 033 0 3 o 6666 608 506 605 10 070 506 07 051 07010509 NNNN N N N N N N N N N N N N N N N N N N N N N N N N N N N N N N N N N N N N N N «u — — — N — r a N N —«w r Mud” Ndm—N N NNNNNNN——-NN—N NNNNN—-dNNNN—NNNNNNNNNNNdN—NNN—NNNNN-fid— 2 02 2 2 22 35321 7115311 2333 709 506 ~05 08 061 608 10 076 08030606 1 I15 0 6 O 9 0 6666 607 503 505 07 026 505 07 039 06010906 u-...6”——_wu__.—~w~—u~—w_-____u-ddwu~_—~u—-— ‘dd‘d-d—d PJ'JPJ'S'PNNO €33 _WNNWNMO ~—————OOOOOOOOOOOOOOOOOO NNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNN-—o—----—-—-----—---—-—----—----—o----------—----—-- OWNNNNOOWPG'PPkMNN O-I-i-d—d—d-‘d-‘OOOOOOOOOOOOOOOOOOO OPWNN—WWNN-‘d €338: Wded-WWNWM‘WWWWW NNNNNNNNNNNNd-‘y—I—l-fl--'—0—d-fi-i--‘-‘I‘dd-fld-‘d-l-fidd-'ddNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNdd—D-i—l-O-i-P-i OOOOOOOO WNNNN—o—o N .0 cu. _- —- N .60 _- —- .0 .0 _- d —n —6 — N N N N N N N N ~ N N N N N N _n .- —o .u _n -ONNN-----o—-o-o—o—N—o—oo-oo-o-o—u—N—o—o-—NoNoNNNONNN—-NONNN-N—NNNNNON—-N ---—-NN-N-o- -‘ -‘ d N N - N N N N N N N N N N N N N —- N N N -' N ... N N N -' N - N N N v ._. -“|_.“A‘I“I N N N .vyvv‘vvv —I-‘-ldN—N-ON-'dN-fi—fi-INN-id-l-‘d-‘NNN-‘d-HN—I-fi-fi-fi-l-IN—NdN-fiN-NNNNNNNN—O—NNNNNNNNNNNNNdNNNd-‘NNN—NN-‘d Haw-”wao-‘O-‘t-m-mdoddNo-‘O-‘d-lm-iM-‘WMWNNNNNNNdMDNC’N PN—N—vN—‘NNWN-‘Wd-‘NO-fl-‘WN I_4 I‘I‘I‘QAI— M 6—0 NNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNN ' — . b—n-a—oP—or-M#fl—r—MN—Nr-Nw-ADN-N\»WNMNNuMQWWWN-NJPWWWNWO—N—uNm — ---N6»N—N0~NNNNN~WWN—Nw—N—NNNNN—Nu—N—rNr—or—J-NN—r—«n-c-NN—rNr—N-cr—NNrNNWNNN NdNNN—d—NN—NNd—NNN——NNNNNN—NNNNNNNN—v NzNNN-flN-fl-ONNNdNNNNN-‘N-IN-'Nd-NNdNNNd”HNNN”Nd-NNNNN—l—ONNNNN-‘NN-fi—NdNNNNNNN-fldNNNN-ONN-flNNN '131 16 1237022zo71 8666 36 02061 6333 3333 070 0£6 036 060000 66 0 5 069 065666 98 06320662059666 030 0 6 035 033000 081 027 086 075000 2 00'61025 036666 033302710 5 035666 6332052‘0 1'061333 296 066 0 6 060333 05; 0 9 055 066333 1 1 1 1 2 072005620722066000 3939065'039'067000 07 “07720652080333 '62 0562053 063666 05 0 0 0g7 050333 1 1 30%:06820 02082666 095 DE: 076 080333 1 7 1°172°“62 21333 11390 721222110000 19911 6 152 168333 0 11 08 12 ; 202:20 32023333 1 1 1 1°: 1°333 60 1 6 125 1 333 03 1 2 120 0 1 0 6 9000 63 160 088 1 1000 2 8 2 2 0 0 8 086 0 2333 21521 2110821 0666 “6220 921 21 3666 £62 135 119 1 6666 0 2 092 125 0 7 0 8000 3603 1 1 Z 2 1202620 920 2000 ‘6 20‘620562069666 261 0382075 066666 0 0 ‘MW m N 0 06 0 6 0 0 0 6 1 3 1 1 i 2 5 333 066 0 5 0 6 061000 2 2 2 161 0 1 6 3902 1 2 962 7 333 02 031 030 023333- 507 606 606 03 508 506 506 10 608 608 607 15 607 606 306 07 “05 507 “05 07 606 303 303 06 606 605 605 07 306 603 602 10 100 201 201 05 605 506 506 11 201 202 606 05 303 303 606 03 303 603 603 06 “05 303 303 05 606 303 303 06 506 ‘05 505 15 605 302 603 06 202 101 202 05 303 303 606 06 608 608 506 06 503 503 503 07 608 506 505 06 710 609 609 10 505 505 505 07 607 508 607 10 710 608 508 09 609 “07.607 06 710 609 609 15 507 506 508 05 507 505 505 09 509 507 608 09 608 608 506 15 509 608 607 06 709 507 608 16 202 202 202 05 605 605 605 07 ‘05 303 303 07 607 305 605 13 203 306 306 05 605 306 206 08 015 507 05 029 06010003 080 606 07 051 07000908 089 710 15 090 10000508 060 605 06 050 00000006 065 306 03 017 01000103 019 307 06 026 06000007 035 206 03 010 02020606 068 309 03 029 10000000 063 303 05 063 00080603 087 607 07 035 05030310 035 605 03 021 08000808 016 606 05 026 09000707 062 603 06 053 05000505 021 507 07 036 06050706 028 302 03 015 07000708 070 605 09 066 05000506 035 303 03 015 00000207 026 208 03 011 09000905 013 606 07 028 05000605 036 506 06 032 05000005 019 506 08 062 05010505 022 508 08 053 08000008 083 508 11 106 07000102 055 606 07 060 06050605 075 506 10 075 07000709 056 608 05 060 10001010 032 607 07 065 07030607 150 710 11 086 05000005 061 608 05 037 08000210 059 506 15 125 07000708 078 507 09 072 07000808 095 509 13 098 03050507 017 506 05 035 08000709 116 710 11 081 10000510 011 202 06 016 10010105 033 605 05 026 05000707 050 201 02 010 00030801 079 303 07 027 00000009 030 207 05 025 06000000 057 505 11 090 09000806 flu—I~—o~u-o-o‘Ndddd—d-—Ndd~ddd-—-ddd~ddduu‘- ~—rr-rrrr-oorrmr~w~n-m~ 8 MNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNN~NNNNNNNN ————~—-——~——~———oooooooooooooo———————-oooooooooo MNNNNNNNNNNNNNNNNNNNNNNNNMNMNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNd-Id—O-i-fi-O-l-‘d-fi-fi-d-I-‘d—‘d twnrsxuuuunuuuh——4WO 88 dS‘WNWWd-WPWH‘OW _‘d-‘NNNNNNNNNNNNNNNNNNNNNNNNNNNNNN-fi--d-fi—O-i-I—n.—0—--—-—'--—I——---—-l-———Idddd—NNNNNNNNNNNNNNNNNN CMDCh-hd-h-43CH3CHDCHDCHDCHDCHDCNDCNDCHDCHDCD UNNNN-fi-‘OMWWJ’WNNNNd— 132 lgzgi‘g 2 0552059'0321051000 306 606 505 08 053 606 05 027 03000605 £225222 2 375'021'026l060000 306 303 303 06 021 303 05 033 10000005 3323221 2 063 0 1 033 037666 506 605 506 11 070 306 96 036 10000510 i0zggzg 2 9 0 3 061 063666 609 608 508 15 106 608 07 050 10050609 3120;! 2 383211610392088666 303 606 606 06 029 605 07 030 08000609 ;1;£§26 2 021 0:0 025 025333 303 305 605 03 026 606 03 026 05030606 3 ‘ '2; 1 036 0 8'0 6 052666 506 603 603 06 018 606 03 016 00000705 g 5 2 $72 0 9 050 063666 507 606 606 06 015 507 06 029 08000008 3 3 2 15 0 0 068 100666 306 605 605 07 032 605 07 039 08020807 1 2 16 21001126 116333 608 506 505 07 050 605 05 030 07000810 1§ 0 3207620 8000 710 508 508 09 076 609 09 076 10000010 7 713310301019666 710 710 710 09 066 710 09 082 10000810 '.6620 9 0 2 0 9000 609 608 503 15 078 608 12 075 08000805 05 069 065 0 666 710 609 609 07 050 609 05 060 08010507 10030 8 061 066000 710 607 607 08 060 609 08 055 10000909 “86 085l0§9:0 5000 710 710 609 05 066 610 06 037 10100510 15 312011§2‘160333 507 507 606 16 056 505 15 080 07010507 5 0 8000 709 507 506 06 066 608 05 062 08000810 dd” N N” d ad 10 0 6 0 091:10OTOE820 8000 507 506 507 12 106 507 06 068 10031009 1005030 0 7‘0 5666 608 607 605 11 096 605 03 016 02050605 653610311301‘ 9000 505 303 606 09 065 606 11 077 08010210 $2330 0 12121 2000 609 509 609 07 056 608 09 086 10000010 13520 720721023666 507 606 506 11 075 605 06 030 07000708 2'é20 52051‘0 333 609 609 509 07 050 608 08 058 08000909 020 0 1 026 030333 201 202 302 02 010 303 03 012 02020206 2836016108‘1063333 506 506 506 07 036 605 05 023 05010505 078206720 62053666 506 505 505 12 090 505 08 056 08000609 61 06820522053333 609 308 308 03 022 608 02 016 10000010 21gkoi9'025'039333 507 606 605 15 105 606 09 059 06060706 d —‘ ”— N N d . MN-‘NNWWNMWN NNNdN—NNNNNNNN—‘NNNNNNN—NNHNN-‘NNN-IdN-‘N—N—zNN—NNN-‘NNN‘NNN dNNN—ONdNNN-‘NNNNNNN—IN—O-‘NNNN N ‘ ‘dd‘dd—dddddd—dN dd NNN-‘NN N 118306910 7 075333 306 303 605 12 057 605 07 062 07020609 76 066 069 062333 202 201 303 06 066 303 06 069 10000006 g6373i62061 067333 303 202 303 09 057 303 07.036 03000306 06 9 056 057000 605 606 606 07 051 605 06 062 06000507 30‘0022 060 065000 606 609 606 15 136 606 07 053 10001010 9 °065 028 056666 609 609 710 11 086 508 06 055 06000103 062 0 3 067 067333 606 303 303 06 035 605 06 030 06000907 3182026fo 8:076666 607 507 507 12 056 507 09 039 07020600 058306010211063666 505 603 603 05.022 605 06 015 09000509 77 0 2 056 061000 606 606 606 07 033 306 07 036 06000000 2 96 071 063 0 6000 709 607 709 09 076 605 06 036 07010709 1 062 062 062 068666 606 506 506 07 067 506 07 065 07060505 1 NNN—ONHNNN NNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNN r N-INNN—I udddd~d~dfldd~dd~d~fldudedddu~dddefid~d N w—mmmttwun————OO\W-————ommm NNOONNOOMOOd—krG‘O‘OOMOFWOOM-‘dPJ'MOPWfrOOOM- d—d-fi-fi-‘OOOO00OOOOOOOOOOOOOOOOOOd-‘d-‘d-‘d-‘dd-‘-‘OOOOOOOOOOOOO NNNNNNNNNNNNNNNNNNNNNNNNNNNNNN—v—-—-—----—I--—--—-—t—-———-—--—-—-—t——-I NN-‘NNN “WNWNWd—NN—P—t-‘NdNNP-‘f-‘N-‘mwMNNNMNNNWdeWNdN-‘N N dNN-‘NNNN-‘NNNNNN-‘NNHNNNNdeNNNdNNNNNd-INNNdNNNNN-‘NNN—N~N-‘NN NNN NNN nun-I—I—d—d—o—o—o—o—‘dd N—NNN ”N N N -ew—ww-~om—M-4aw—o—m-«onsn-mNr—om—o— N '05 N N N N N wwv —v~' w a: «a no 032-on W;O——WNONm—ONNN~NMNNN—szndu—N—no_o—__.o.‘m_.~_m_.o_m_m_.w d NN-‘NNNN—HNN-fiHNNNNNNNNdNNNNN-‘NNdN-IN-IN-iNNN-IN—‘d-IN-i-l—INd-‘d-lN—ON u nAIA-‘QAIAIAI‘IA‘IAL--A‘IAI‘ . - v - - . v w N N NNNNNNNNd—NNNNN-‘d-‘NNNdN-‘NN-fi N NNNNNNNNNNNNNNNNNNNNNNNN~NNNN 1133 053 050 036 066333 087 065 039 023666 207 165 166 122000 160 058 080 0 2666 112 069 098 0 9666 089 069 106 038000 110 103 053 0§8666 207 129 116 1$0666 092 068 083 0 1000 106 070 060 071333 082 109 067 036000 231 160 161 136000 126 097 123 116666 036 037 031 0 6000 160 162 179 153666 101 026 036 056333 030 060 025 031666 097 108 076 093000 202 078 075 118333 080 066 166 089333 123 086 076 096333 062 021 060 036333 061 028 027 032000 107 068 070 081666 116 126 096 112000 161 071 069 093666 062 052 032 062000 052 056 099 069000 709 606 606 507 505 505 710 710 709 709 608 608 709 505 505 608 506 507 608 505 606 608 506 505 606 202 202 609 508 806 709 507 505 507 507 507 605 605 605 605 605 605 206 305 306 605 605 605 306 306 306 507 605 606 506 506 506 303 303 201 307 ‘05 “07 303 303 ‘0‘ 606 606 606 606 606 506 603 709 709 609 507 507 606 606 605 606 606 606 11 059 505 10 06% 07030708 12 06k 506 08 081 09010705 10 05 06 07 09 08 05 07 06 10 09 05 11 09 07 15 09 06 06 06 07 10 15 10 10 095 709 09 080 10000000 033 709 06 058 09000810 037 507 07 060 08000709 050 507 07 Obh 08000000 039 805 07 050 09000810 055 506 11 080 08050508 018 202 05 027 08000108 029 508 15 063 07000005 031 608 06 031 08020608 076 508 13 101 10020707 039 505 10 057 07010509 028 E06 05 031 07010506 606 16 102 10021010 506 16 098 05000005 505 03 015 06000306 303 07 016 05010605 605 09 068 07000608 605 05 032 08010508 606 03 019 08000808 607 O7 061 08000808 606 06 012 03060106 606 07 067 07000000 150 607 10 088 07030810 08k 508 09 058 08010809 205 03 028 09000000 05 039 £07 07 063 10001010 ‘—‘~‘—udd~—_d--—~dd—-d~udd MICHIGAN STRTE UNIV. LIBRRRIES \IHI llllll I" llll 1| MN I!" ll llll VI 111 H 11 1|” 1111 MINI 31293008757779