What Motivates Children's Behavior and Emotion? Joint...

11
Journal of Personality and Social Psychology 1993, Vol. 65, No. 4, 781-791 Copyrighi 1993 by the American Psychological Association, Inc. 0022-3514/93/S3.00 What Motivates Children's Behavior and Emotion? Joint Effects of Perceived Control and Autonomy in the Academic Domain Brian C. Patrick, Ellen A. Skinner, and James P. Connell This study examined the contribution of perceived control and autonomy to children's self-reported behavior and emotion in the classroom (JV = 246 children ages 8-10 years). Multiple regression analyses revealed unique effects of autonomy over and above the strong effects of perceived control. In addition, both sets of perceptions (and their interaction) were found to distinguish children who were active but emotionally disaffected from those who were active and emotionally positive. Specific predictions were also tested regarding the effects of (a) control attributions to 5 causes and (b) 4 reasons for task involvement that differed in degree of autonomy on children's active (vs. passive) behavior and 4 kinds of emotions: boredom, distress, anger, and positive emotions. Impli- cations of thefindingsfor theories of children's motivation are discussed, as well as for diagnostic strategies to identify children at risk for motivational problems. Perceived control is a powerful influence on behavior and emotion. Research from a wide variety of perspectives has docu- mented this connection over the last three decades. Attribu- tions of success and failure to causes that differ on a variety of dimensions predict exertion, persistence, and emotional reac- tions to performances, such as pride, shame, and anger (Weiner, 1979,1985). Learned helplessness studies have shown that indi- viduals who feel that there is no connection between their be- haviors and desired outcomes are passive, depressed, and do not perform very well (Abramson, Seligman, & Teasdale, 1978). Research on self-efficacy reveals that people who do not believe that they can produce effective responses become anxious and fearful, and retreat in the face of challenge (Bandura, 1986). Constructs and findings from major theories have been inte- grated in a new conceptualization of perceived control (Con- nell, 1985; Skinner, 1991; Skinner, Chapman, & Baltes, 1988; Skinner & Connell, 1986). This conceptualization distin- guishes between children's beliefs about (a) "what it takes to do well," or the causes of success and failure and (b) "whether I've got it," or the extent to which the self has access to particular Brian C. Patrick, Ellen A. Skinner, and James P. Connell, Depart- ment of Psychology and Graduate School of Education and Human Development, University of Rochester. This research has been supported by grants from the W T. Grant Foundation, National Institute of Mental Health Training Grant 527594, and National Institute of Child Health and Human Develop- ment Research Grant HD19914 and by the Jacob K. Javits Fellows Program. We acknowledge the valuable contributions of our colleagues in the Motivation Research Group, Edward Deci and Richard Ryan. We also thank the other members of the research project who helped collect these data, Michael Belmont, Jeff Altman, and Peter Usinger. We thank the children, parents, teachers, and administrators in the school district where this research was conducted. Correspondence concerning this article should be addressed to El- len A. Skinner, who is now at the Department of Psychology, Portland State University, Box 751, Portland, Oregon 97207-0751. causes (Skinner et al, 1988; Skinner, Wellborn, & Connell, 1990). Referred to as strategy and capacity beliefs, 1 respectively, combinations of these beliefs have proved useful for identifying profiles of perceived control that promote and undermine chil- dren's motivation, behavior, and emotion. For example, in the academic domain, a clear picture emerges (Skinner et al, 1990). The children most actively en- gaged in the classroom are those who believe (a) effort is an important cause of school success and failure, and they them- selves can exert effort (high effort strategy and capacity beliefs); (b) that although ability is not necessary for success, they them- selves are smart (low ability strategy beliefs and high ability capacity beliefs); and (c) that they have access to powerful others and are lucky (high powerful others and luck capacity beliefs). In contrast, children who are most disaffected from school ac- tivities believe (a) they are incapable of exerting effort and are not smart (low effort and ability capacity beliefs); (b) that power- ful others and luck are needed to succeed, but they themselves cannot influence others and are unlucky (high strategy and low capacity for powerful others and luck); and (c) that they don't know the causes of success and failure in school (high unknown strategy beliefs). Pairs of strategy and capacity beliefs for each of the five causes examined in the academic domain (effort, ability, powerful others, luck, and unknown) are strong predic- tors of children's behavior and emotion in the classroom. Taken together, profiles of beliefs predicted to promote and under- mine children's motivation account for more than 25% of the variance in teacher's ratings of children's engagement and disaf- fection in the classroom (Connell & Wellborn, 1991; Skinner, 1991; Skinner etal, 1990). Surprisingly, the empirical success enjoyed by control con- structs has had an unintended side effect for the study of moti- vation. Because perceived control makes such a strong contri- bution to behavior and emotion, enthusiasm for the search for 1 These were originally referred to as means-ends beliefs and agency beliefs (Skinner, Chapman, & Baltes, 1988). 781

Transcript of What Motivates Children's Behavior and Emotion? Joint...

Journal of Personality and Social Psychology1993, Vol. 65, No. 4, 781-791

Copyrighi 1993 by the American Psychological Association, Inc.0022-3514/93/S3.00

What Motivates Children's Behavior and Emotion? Joint Effects of PerceivedControl and Autonomy in the Academic Domain

Brian C. Patrick, Ellen A. Skinner, and James P. Connell

This study examined the contribution of perceived control and autonomy to children's self-reportedbehavior and emotion in the classroom (JV = 246 children ages 8-10 years). Multiple regressionanalyses revealed unique effects of autonomy over and above the strong effects of perceived control.In addition, both sets of perceptions (and their interaction) were found to distinguish children whowere active but emotionally disaffected from those who were active and emotionally positive.Specific predictions were also tested regarding the effects of (a) control attributions to 5 causes and(b) 4 reasons for task involvement that differed in degree of autonomy on children's active (vs.passive) behavior and 4 kinds of emotions: boredom, distress, anger, and positive emotions. Impli-cations of the findings for theories of children's motivation are discussed, as well as for diagnosticstrategies to identify children at risk for motivational problems.

Perceived control is a powerful influence on behavior andemotion. Research from a wide variety of perspectives has docu-mented this connection over the last three decades. Attribu-tions of success and failure to causes that differ on a variety ofdimensions predict exertion, persistence, and emotional reac-tions to performances, such as pride, shame, and anger (Weiner,1979,1985). Learned helplessness studies have shown that indi-viduals who feel that there is no connection between their be-haviors and desired outcomes are passive, depressed, and donot perform very well (Abramson, Seligman, & Teasdale, 1978).Research on self-efficacy reveals that people who do not believethat they can produce effective responses become anxious andfearful, and retreat in the face of challenge (Bandura, 1986).

Constructs and findings from major theories have been inte-grated in a new conceptualization of perceived control (Con-nell, 1985; Skinner, 1991; Skinner, Chapman, & Baltes, 1988;Skinner & Connell, 1986). This conceptualization distin-guishes between children's beliefs about (a) "what it takes to dowell," or the causes of success and failure and (b) "whether I'vegot it," or the extent to which the self has access to particular

Brian C. Patrick, Ellen A. Skinner, and James P. Connell, Depart-ment of Psychology and Graduate School of Education and HumanDevelopment, University of Rochester.

This research has been supported by grants from the W T. GrantFoundation, National Institute of Mental Health Training Grant527594, and National Institute of Child Health and Human Develop-ment Research Grant HD19914 and by the Jacob K. Javits FellowsProgram.

We acknowledge the valuable contributions of our colleagues in theMotivation Research Group, Edward Deci and Richard Ryan. We alsothank the other members of the research project who helped collectthese data, Michael Belmont, Jeff Altman, and Peter Usinger. Wethank the children, parents, teachers, and administrators in the schooldistrict where this research was conducted.

Correspondence concerning this article should be addressed to El-len A. Skinner, who is now at the Department of Psychology, PortlandState University, Box 751, Portland, Oregon 97207-0751.

causes (Skinner et al, 1988; Skinner, Wellborn, & Connell,1990). Referred to as strategy and capacity beliefs,1 respectively,combinations of these beliefs have proved useful for identifyingprofiles of perceived control that promote and undermine chil-dren's motivation, behavior, and emotion.

For example, in the academic domain, a clear pictureemerges (Skinner et al, 1990). The children most actively en-gaged in the classroom are those who believe (a) effort is animportant cause of school success and failure, and they them-selves can exert effort (high effort strategy and capacity beliefs);(b) that although ability is not necessary for success, they them-selves are smart (low ability strategy beliefs and high abilitycapacity beliefs); and (c) that they have access to powerful othersand are lucky (high powerful others and luck capacity beliefs).In contrast, children who are most disaffected from school ac-tivities believe (a) they are incapable of exerting effort and arenot smart (low effort and ability capacity beliefs); (b) that power-ful others and luck are needed to succeed, but they themselvescannot influence others and are unlucky (high strategy and lowcapacity for powerful others and luck); and (c) that they don'tknow the causes of success and failure in school (high unknownstrategy beliefs). Pairs of strategy and capacity beliefs for eachof the five causes examined in the academic domain (effort,ability, powerful others, luck, and unknown) are strong predic-tors of children's behavior and emotion in the classroom. Takentogether, profiles of beliefs predicted to promote and under-mine children's motivation account for more than 25% of thevariance in teacher's ratings of children's engagement and disaf-fection in the classroom (Connell & Wellborn, 1991; Skinner,1991; Skinner etal, 1990).

Surprisingly, the empirical success enjoyed by control con-structs has had an unintended side effect for the study of moti-vation. Because perceived control makes such a strong contri-bution to behavior and emotion, enthusiasm for the search for

1 These were originally referred to as means-ends beliefs and agencybeliefs (Skinner, Chapman, & Baltes, 1988).

781

782 B. PATRICK, E. SKINNER, AND J. CONNELL

other influential constructs has been dampened. When the re-cent history of motivation is told, theories of perceived controldominate the field (Weiner, 1990). Nevertheless, theorists andresearchers in this area realize that sources of motivation areoperating that are outside the boundaries of current conceptual-izations of perceived control. What maintains behavioral in-volvement if perceived control falters in the face of challengesor failure? What guides the selection among tasks that are per-ceived as equally controllable? What initiates behavior whenexperiences with an activity are too brief to result in reliableestimates of control? Why do some children who are competentapproach learning activities with enthusiasm and interest andother competent children approach them with boredom or anx-iety?

Autonomy Versus Control

One explanation focuses on children's autonomy, denned as"a sense of being choiceful in one's actions and experiencingoneself as the locus of initiation of those actions" (Connell &Ryan, 1987, p. 5; see also Deci & Ryan, 1985; DeCharms, 1976;White, 1959). Before the joint and synergistic effects of per-ceived control and autonomy can be studied, however, two con-ceptual questions must be addressed. First, is autonomy differ-ent from control? And, second, does autonomy influence chil-dren's behavior and emotion?

Consistent with most people who study autonomy, broadlydenned, we argue that control and autonomy are conceptuallydistinct (see DeCharms, 1981; also Connell & Wellborn, 1991;Deci & Ryan, 1985; Dweck & Leggett, 1988; Harter, 1981; Ni-cholls, 1984; Skinner & Connell, 1986). Control refers to theconnection between behaviors and outcomes; it is the extent towhich a person feels capable of producing desired and prevent-ing undesired events; the opposite of control is helplessness. Au-tonomy refers to the connection between volition and action; itis the extent to which a person feels free to show the behaviors ofhis choice; nonautonomous behaviors include both complianceand defiance, which have in common that they are reactions toothers' agendas and not freely chosen.

It is easy to imagine situations of high control and low auton-omy. Many would argue that schools are prototypical in thisregard. In schools, contingencies are extremely high and consis-tent. Children understand the actions that lead to success andfailure. However, in many classrooms, even successful childrenare not free to choose their own behaviors or to follow their owninterests. Conformity (and hence loss of autonomy) is the behav-ior required to control desired outcomes. In fact, it is possible toargue that, in general, extremely high behavioral contingenciesfor desired outcomes make it difficult to exercise autonomy;the pressure created by the very contingencies that provide con-trol decrease the individual's psychological freedom not to en-gage in the behaviors (Deci & Ryan, 1985).

More controversial is the possibility that high autonomy canexist in the absence of control. Some theories, for example,self-determination theory, argue that high competence (andhence control) is a prerequisite for autonomous behavior (Deci& Ryan, 1985). If no control is possible, a state of amotivationexists, no matter how much freedom is also present. A counter-

argument could be made for cases in which individuals are notconcerned with the production of outcomes. Even in schools,children can freely and choicefully participate in some activi-ties (such as art or music), with total disregard for the success orfailure of the final product.2

Regardless of the position one takes on this issue in general, itis an empirical question about whether the influence of controland the influence of autonomy on behavior and emotion in aspecific domain and at a specific age are redundant, additive, orinteractive. That question was at the heart of the present study.Before we describe the hypotheses, however, we outline theconceptualization of autonomy used in the study.

Autonomy and Control

Autonomy

One manifestation of autonomy is found in individual's rea-sons for engaging in activities. A continuum of autonomy canbe used to order qualitatively different reasons for involvement(Chandler & Connell, 1987; Connell & Ryan, 1984; Deci &Ryan, 1985; Ryan & Connell, 1989). The least autonomous rea-sons for initiating behavior are external, or originating totallyoutside the person, and include such motivators as expectationsof rewards, threats, bribes, and punishments. In these cases,people do not feel free to show the behaviors they choose forfear of the consequences. Next along the continuum of auton-omy are reasons that, although internalized, retain their evalua-tive and pressured character. These are called introjected rea-sons. In this case, an individual is not free to show desiredbehaviors for fear of what she will think about herself (e.g., thatshe is a bad person).

The next reasons, referred to as identified reasons, althoughnot purely intrinsic, are more autonomous than external or in-trojected reasons because they are freely chosen and individu-ally endorsed. Identified reasons include motivators that wereoriginally external but have been internalized into the personal

2 One difficulty in understanding the distinction between auton-omy and control is the result of an historical accident in terminology.Because control and autonomy have independent theoretical tradi-tions (DeCharms, 1981), no attempt has been made to use labels sys-tematically. Both are concerned with the "location" or "locus" of theinfluence, in one case, the influence on outcomes (locus of control),and in the other, the influence on behavior (locus of causality). Andboth have tried to distinguish internal from external influences. How-ever, cumulative research has shown that the distinction between in-ternal vs. external is not very useful in either control or autonomy.Many internal causes are not controllable (e.g., ability). And many ac-tions that originate inside the skin are not choiceful or self-determined(e.g., compulsions). To make matters more complicated, researchersdiscussing the kinds of contexts that undermine autonomy have re-ferred to them as controlling (vs. informational; Deci & Ryan, 1985;Ryan, 1982). To avoid these terminological confusions, we use controlonly in reference to perceived control and refer to contexts which un-dermine autonomy as coercive. We do not use locus or internal vs. exter-nal to refer to either control or autonomy. Otherwise, we try to use theconstruct labels in a way consistent with their typical use in the respec-tive areas.

PERCEIVED CONTROL AND AUTONOMY 783

value system. In this case, the person engages in the behaviorbecause it is personally important to him or her. Next are inte-grated reasons that, although not seen in children, refer to flexi-ble, balanced interrelations among multiple, competing identi-fied goals from different domains. Last are intrinsic reasons forinvolvement in an activity. Purely intrinsic motives focus oninvolvement in the task for its own sake, with emphasis on theprocesses and not the outcomes of the activity. They are charac-terized by enthusiasm, spontaneity, excitement, intense con-centration, rapt attention, complete involvement, and joy.

In the academic domain and with children, research hasbeen focused on how children with different reasons for involve-ment in school activities, such as studying for tests or complet-ing homework, also show different patterns of classroom en-gagement and performance. For less autonomous children, per-formances are marked by fear, tension, pressure, defensivestrategies, and attempts to evade the activity. In contrast, formore autonomous children, performance in the face of obsta-cles is characterized by exertion, persistence, strategy testing,and optimism (Connell & Wellborn, 1991; Deci & Ryan, 1985;Dweck & Leggett, 1988; Nicholls, 1984).

This perspective provides a basis for predictions about auton-omy as a powerful motivator of performance and affect, inde-pendent of perceived control (Connell & Ryan, 1984; Ryan &Connell, 1989; Skinner & Connell, 1986). It elaborates multiplereasons for regulating behavior and postulates that they differ-entially influence motivation because of the extent to whichthey differ on autonomy. Taken together with theories of per-ceived control, this perspective was used to derive the followingpredictions about the joint effects of control and autonomy onchildren's behavior and emotion during learning activities.

Hypotheses About Control and Autonomy

Behavior. Both perceived control and autonomy have beenshown to predict to exertion, attention, persistence in the faceof difficulties, and sustained involvement in learning activities(Abramson et al., 1978; Bandura, 1986; Connell & Ryan, 1984;Connell & Wellborn, 1991; Deci & Ryan, 1985; Ryan & Con-nell, 1989; Weiner, 1985). Hence, we predicted that both per-ceived control and autonomy would show individual relationsto children's behavioral engagement in the classroom. We didpredict, however, that perceived control would show a strongerrelation to behavior than autonomy. We reasoned that, al-though higher control and more autonomy would both predictto more active behavior, the effects of low control and low au-tonomy would differ. Children who have low perceived controlwould not even try, and as a result, effortful behavior would betotally absent (which Deci & Ryan, 1985, described as amotiva-tion; Skinner, Wellborn, & Connell, 1990). In contrast, childrenwith low perceived autonomy might nevertheless show mini-mal behavioral compliance (e.g., because they are motivated byexternal reasons, such as fear of punishment).

We also tested two new hypotheses. First, we expected thatperceived control and autonomy, representing separate sourcesof motivation, would contribute uniquely to motivated behav-ior. Although it might be expected that perceived control, as amajor predictor of effortful performance, would overshadow

autonomy, we predicted that, even given the large portion of thevariance in behavior for which perceived control can account,autonomy would account for additional unique variance (Con-nell & Ryan, 1984; Skinner & Connell, 1986). Second, we pre-dicted an interaction between perceived control and autonomy,such that the highest levels of behavioral involvement would beshown by children with high perceived control and high auton-omy (Connell & Wellborn, 1991).

Emotion. As with behavior, we predicted that both controland autonomy would make unique and interactive contribu-tions to children's emotion during learning activities. However,more differentiated predictions were also made about fourkinds of emotions that children experience during learning ac-tivities: positive emotions (interested, happy, and relaxed), bore-dom, distress, and anger.3 As predictors, we considered attribu-tions to five causes (effort, ability, powerful others, luck, andunknown causes) and four reasons reported by children (exter-nal, introjected, identified, and intrinsic).4

First, we hypothesized that children who are more intrinsi-cally motivated and rely more on effort would express morepositive emotions in learning situations. These two kinds ofmotivation have been shown to have the most beneficial effectson emotion. Second, children who were bored were expected tobe very low in intrinsic motivation. This link was expected tobe especially strong, because interest is the hallmark of intrin-sic motivation. Third, children who feel distressed would be-lieve that they lack the ability that is necessary for success. Lossof control due to internal, stable, and global causes is consid-ered a risk factor for sadness or depression (Abramson, Selig-man, & Teasdale, 1978; Nolen-Hoeksma, Girgus, & Seligman,1986). Fourth, children who are more distressed when involvedin learning tasks would be more introjected in their reasons;anxiety would stem from the pressures such children exert onthemselves. In addition, distressed children would think thatschool performance is dependent on luck or unknown factors;in this case, anxiety would result from the uncertainty con-nected with these causes. Finally, we predicted that angerwould be the result of external reasons for task involvement andthe conviction that powerful others are needed for school suc-cess. We expected that external reasons would be an especiallystrong predictor of anger; anger and resentment would be theresult of being pressured by others.

Behavior and emotion. In general, behavior and emotionare positively correlated. When one feels anxious or sad, onewithdraws from the situation; when one is interested, one tries

3 Originally, we made predictions about the emotions as bipolar con-structs, that is, happiness-sadness, relaxation-anxiety, interest-bore-dom, and anger. The predictions presented here are based on the re-sults of the factor analysis of our measure of children's emotion in theclassroom (see Method section). We appreciate the suggestion of ananonymous reviewer that it was too confusing to include predictionsmade before the factor analysis. Because this change made no differ-ence to our data analyses or interpretations, we followed this sugges-tion.

4 Because integrated reasons for involvement are not seen in children(Chandler & Connell, 1987), that level of autonomy was not included inthe current study.

784 B. PATRICK, E. SKINNER, AND J. CONNELL

harder. However, we were concerned about children for whombehavior and emotion become uncoupled, specifically, thosechildren who continue to show behavioral involvement evenwhen they are emotionally disaffected. These children are ofspecial interest because they are not likely to receive the helpthey need. Anxious or bored children are not typical targets ofinterventions as long as they continue to show behavioral in-volvement in classroom activities.

We reasoned that perceived control was not likely to be veryinformative about these children. Perceived control operatesthe same way on behavior and emotion: High perceived controlleads to high involvement and enthusiastic emotions; whencontrol is low, children feel bad and stop working. We thoughtthat autonomy might be the key to understanding children forwhom behavioral activity continues even when emotional en-ergy is gone. Children who are feeling bored or anxious maynevertheless maintain behaviors if they are being externally orinternally pressured.

Hence, we formulated two hypotheses about combinations ofbehavior and emotion. First, we predicted that, because lack ofcontrol is effective in stopping behavior, the primary differencebetween children who are unhappy and passive (low emotionand low behavior) and those who are unhappy but active (lowemotion and high behavior) would be their perceived control.Second, because low autonomy can maintain behavior even inthe absence of enthusiasm, we expected that the primary dif-ference between children who are unhappy but active (highemotion and high behavior) and those who are active and happy(high behavior and high emotion) would be their autonomy.

Method

Sample and Procedures

Participants were 264 children from two elementary schools in asuburban and rural district in upper New York state. Children wereequally divided by sex and grade from Grades 3,4, and 5 (M age = 8.3years, SD = .45 years). The school district was middle to lower middleclass with a small minority population. Children completed the assess-ments used for this study as part of a larger research project. Assess-ments were conducted in three 45-min sessions in the classroom bytwo trained administrators without the teacher present. Children an-swered questions at their seats while one administrator read the ques-tions aloud and one circulated in the classroom to answer any ques-tions and to make sure that children were keeping up.

Measures

Three assessments were used for this study: perceived control, au-tonomy, and engagement in the classroom. All items were answered ona 4-point scale from not at all true for me (1) to very true for me (4). Allscores were calculated by averaging the items for the respective scaleand so could range from ! to 4.

Perceived control. The Student Perceptions of Control Question-naire (SPOCQ, Wellborn, Connell, & Skinner, 1989; Skinner et al.,1988) was used to tap children's strategy and capacity beliefs for fivecauses: effort, ability, powerful others, luck, and unknown causes. Allscales consisted of six items, equally divided into success and failure.The strategy beliefs scales tapped children's beliefs about the extent towhich each cause is a necessary condition for producing success and

preventing failure in school (e.g., effort: "If 1 want to do well on myschoolwork, I just need to try hard" and ability: "If I'm not smart, Iwon't get good grades"). For the five strategy belief scales, internalconsistencies ranged from .65 to .82, M a = .73. The correspondingcapacity beliefs scales tapped children's beliefs about the extent towhich they possessed or had access to the four known causes (e.g.,effort: "I can't seem to try very hard in school" and ability: "When itcomes to school, I'm pretty smart"). For the four capacity belief scales,«s = .59-.79, Ma = .70.

For purposes of this study. Strategy X Capacity Beliefs interactionscores for each of the four known causes were constructed. For effortand ability interaction scores, a high score indicates an attribution ofsuccess to that cause and a low score an attribution of failure. Interac-tion scores for ability were calculated by multiplying the additive in-verse of strategy beliefs by capacity beliefs (i.e., low strategy and highcapacity beliefs: "Ability is not essential but I'm smart"). For effort, theinteraction score was calculated for success by multiplying strategy bycapacity beliefs and for failure by multiplying strategy beliefs by theadditive inverse of capacity beliefs, and then these two effort scoreswere subtracted from each other. Hence, high scores indicated higheffort strategy and capacity beliefs ("Effort is important and I can tryhard"), whereas low scores indicated high effort strategy and low effortcapacity beliefs ("Effort is important but I am unable to try hard").Finally, for powerful others and luck, a high score indicates an attribu-tion of failure to that cause and a low score indicates an attribution ofsuccess. Interaction scores for powerful others and luck were calculatedby multiplying strategy beliefs by the additive inverse of capacity be-liefs (i.e., high strategy and low capacity beliefs: "Luck is an importantcause and I'm unlucky").

In addition, a maximum control score (ConMax) was calculated bycombining profiles of beliefs. Higher scores on this profile indicatehigh strategy beliefs for effort, and low strategy beliefs for ability, aswell as high capacity beliefs for all four causes. Low scores on thisprofile indicate high strategy beliefs for powerful others, luck, andunknown as well as low capacity beliefs for all four causes (Skinner etal., 1990). In sum, six perceived control scores were used in this study:interaction scores for effort, ability, powerful others, and luck; un-known strategy beliefs; and maximum control.

Autonomy. Children's perceived autonomy was assessed using theSelf-Regulatory Style Questionnaire (SRQ, Connell & Ryan, 1984;Ryan & Connell, 1989). The SRQ consists of four scales that differ onthe extent to which reasons for task involvement are autonomous. Chil-dren rate each set of items as answers to questions about why theyperform school activities (e.g., "Why do 1 do my homework?"). Thesix-item External scale (a = .78) taps reasons outside the child thatimply pressure from external sources (e.g., "because I'll get in trouble ifI don't"). The five-item Introject scale (a = .75) taps reasons that areinternalized but directly incorporated from external reasons and so arestill pressured (e.g., "because I'll feel bad about myself if I don't do it").The four-item Identified scale (a = .61) taps reasons that are internal-ized into the value system of the child and so are choiceful and unpres-sured (e.g., "because I want to learn new things"). The six-item Intrinsicscale (a = .85) taps reasons of pure fun and enjoyment (e.g., "because it'sfun").

These four scales have been shown to be factorially distinct andvalid as indicatorsof the amount of autonomy with which school activi-ties are carried out (Ryan & Connell, 1989). A summary score, referredto as the Relative Autonomy Index (RAI) was also calculated in whichthe four scales were weighted according to their autonomy (- 3, - 1 , +1,and +3, for external, introject, identified, and intrinsic, respectively)and summed. Hence, five autonomy scores were used in the study:external, introjected, identified, intrinsic self-regulation (higherscores on which indicated more of the respective reasons), and the

PERCEIVED CONTROL AND AUTONOMY 785

RAI, higher scores on which indicated more autonomy in initiatinglearning activities.

Behavior and emotion. Children's motivated behavior in theclassroom was assessed using an 18-item self-report scale (a = .87),which tapped children's effort, persistence, attention, and participa-tion during the initiation and completion of learning activities (Well-born, 1991; Wellborn & Connell, 1987). The scale was equallyweighted by positive and negative items (e.g., "The first time myteacher talks about a new topic, I listen carefully"; "When I'm in class,I just act like I'm working"). These items were factorially distinct fromthe emotion items. Children's reports of their behavior in theclassroom have been found to be significantly related to teachers' re-ports of these same behaviors (Skinner & Belmont, in press). Hence,we treated these self-reports as indicators of children's actual behav-ioral involvement in class.

The scores used to reflect children's emotions in the classroom werethe result of exploratory factor analyses on a set of items designed to tapa range of emotions experienced during the initiation and completionof learning activities. Children rated 35 items for the extent to whichthey corresponded to emotions they felt in school. Three stems wereused ("When we start something new in class, I feel. . ."; "When I'mworking in class, I feel . . ."; and "When I'm in class I feel . . .").Twenty negative emotion words were used to tap five dimensions:bored (e.g., tired, bored, and sleepy), worried (e.g., scared, nervous, andworried), sad (e.g., sad and unhappy), bad (e.g., bad and terrible), andangry (e.g., mad and angry). Sixteen positive emotions were used torepresent the polar opposites of four of the negative dimensions: inter-ested (e.g., interested and involved), relaxed (e.g., relaxed and comfort-able), and happy (e.g., happy and glad).

After removing items on the basis of psychometric criteria, such aslow internal consistencies or lack of correlation with synonyms, weentered the remaining 20 items (12 negative and 8 positive) in an explor-atory factor analysis using oblique (Promax) rotation. We expected fourfactors (three of which would be bipolar positive-negative) correspond-ing to the four dimensions used to generate items. The actual factorpattern differed from that expected in two major ways. First, littledifferentiation of the positive items was found; in general, positiveitems all loaded on a single factor and did not load on the same factorsas their designated negative items. When separate factor analyses weresubsequently performed on the positive and negative items alone, thepositive items were accounted for by a single factor (a = .88). Second,for the negative items, three instead of four factors emerged: bored,worried-sad, and angry. As a result, three negative emotion scoreswere used: bored (4 items, a = .76), distress (worried-sad, 6 items, a =.75), and angry (2 items, interitem correlation = .77). In sum, for behav-ior and emotion indicators, the following scores were used: motivatedbehavior; positive emotion; bored, distressed, and angry emotions;and total emotion (average of positive and negative emotion scores).

Results

Data analyses were designed to answer three questions: (a)Does autonomy contribute uniquely to the primary dependentvariables, behavior and emotion, over and above the contribu-tion of perceived control? (b) How do specific attributions andreasons relate to specific emotions? and (c) Do perceived controland autonomy differentially characterize children with differ-ent motivational profiles, or combinations of behavior (active orpassive) and emotion (positive or negative)? As a context forinterpreting the analyses that address these questions, we pres-ent information regarding mean levels and intraconstruct andinterconstruct correlations.

Table 1Means and Standard Deviations for the Measures of Behavior,Emotion, Perceived Control, and Autonomy

Variable M SD

BehaviorEmotion

PositiveBoredomDistressAnger

Perceived Control (ConMax)"Effort"Ability"Powerful others"Luck"Unknown

Autonomy (RAI)a

ExternalIntrojectedIdentifiedIntrinsic

3.283.102.931.931.641.92

31.078.558.382.633.351.911.142.772.623.532.84

0.490.540.640.790.610.90

15.841.332.771.911.880.684.280.740.780.500.82

Note. N = 264 children, Grades 3-5. Scores could range from 1 to 4,except where indicated. ConMax = maximum control profile; RAI =Relative Autonomy Index.a Composite score; formula detailed in Method section."Strategy X Capacity interaction scores; scores could range from1 to 16.

Initial Analyses

Descriptive statistics. Table 1 presents means and standarddeviations for both the overall composite variables and the sub-components of behavior, emotion, autonomy, and perceivedcontrol. The range of variance, as indicated by the standarddeviations, was satisfactory, revealing comparable variabilityacross constructs. For one subscale, the mean fell within onestandard deviation of the maximum possible value: the identi-fied subscale of autonomy. Given the indication of potentialceiling effects, correlations involving this variable might be at-tenuated.

Intraconstruct correlations. The correlations among sub-scales that make up each construct are presented in Table 2. Foremotion, the positive subscale was highly negatively correlatedwith boredom. The three negative emotion subscales (bore-dom, distress, and anger) were moderately intercorrelated(range = .32-47). The interdependence among the subscales forthe dependent variables might make it difficult to find supportfor the differentiated predictions.

The four subscales of the autonomy measure were intercorre-lated with each other in the simplex pattern of ordered correla-tions reported by Ryan and Connell (1989). Because these sub-scales are ordered according to a continuum of autonomy (fromexternal to introjected to identified to intrinsic), Ryan and Con-nell (1989) hypothesized and subsequently demonstrated thatsubscales that are closer to each other on this continuum aremore highly correlated. Our data differed from this simplexpattern primarily in the correlations involving identified rea-sons: between the introjected and identified subscales (r = .09for our data vs. r = .47, averaged across four samples for Ryan &

786 B. PATRICK, E. SKINNER, AND J. CONNELL

Table 2Intmconstruct Correlations for Emotion,Perceived Control, and Autonomy

Variable 1

Emotion

1. Positive2. Boredom3. Distress4. Anger

1. Effort2. Ability3. Powerful others4. Luck5. Unknown

1. External2. Introjected3. Identified4. Intrinsic

—-.59-.49-.18

—.55.32

Perceived control

.51-.55-.62-.50

—-.47-.52-.39

Autonomy

.55-.10-.27

—.09

-.05

—.48

—.55.36

—.65

.52 —

Note. N= 264 children, Grades 3-5. Behavior is not included becauseit is not broken into subcomponents. Coefficients greater than .21 aresignificant at the .001 level, greater than. 16 at the .01 level, and greaterthan. 13 at the .05 level.

Connell, 1989) and between the identified and intrinsic sub-scales (r = .62 for our data vs. r = .46, from Ryan & Connell,1989). Both the lower identified-introjected correlation andthe higher identified-intrinsic correlation can be explained bythe slightly different flavor of the identified items used in thecurrent study, which included items such as "Why do I try to dowell in school? Because I enjoy doing schoolwork well" (Well-born & Connell, 1987). Including words like enjoy, in effect,increased the conceptual similarity of this subscale to the in-trinsic subscale, while decreasing the similarity to the intro-jected subscale.

The five control subscales (effort, ability, powerful others,and luck strategy by capacity beliefs and unknown strategybeliefs) were moderately intercorrelated. Correlations rangedfrom .35 (powerful others and unknown) to —.62 (effort andluck). The interdependence among the subscales again high-lighted the importance of examining the unique effects of theseindependent variables on behavior and emotion in subsequentanalyses.

Interconstruct correlations. Tables 3, 4, and 5 contain thezero-order correlations between constructs, first for the com-posite variables and then separately by subscale. The behaviorand emotion constructs were closely interrelated (see Table 3),indicating about 50% shared variance between the two primarydependent variables. Especially interesting was the degree ofinterrelatedness between the two independent variables, auton-omy (RAI) and perceived control (ConMax). As can be seen,this relationship was of a somewhat lesser but still considerablemagnitude, again raising the issue of the unique predictivevalue of these independent variables. Later analyses explored

Table 3Correlations Among Composite Measures of Behavior,Emotion, Perceived Control, and Autonomy

Measure 1

1.2.3.4.

BehaviorEmotionPerceived Control (ConMax)Autonomy (RAI)

—.70.69.54

.67

.55 .52 —

Note. N= 264 children, Grades 3-5. All coefficients significant at the.001 level. RAI = Relative Autonomy Index; ConMax = maximumcontrol profile.

this issue in depth. The pattern of correlations suggested thatthe perceived control composite was more closely related toboth behavior and emotion than was the autonomy composite.Although the RAI was significantly related to both (Mr- .55),the zero-order correlations of ConMax to the dependent vari-ables were somewhat higher (M r = .68).

Unique Effects of Perceived Control and Autonomy onBehavior and Emotion

To determine the unique effects of both perceived controland autonomy on the dependent variable composites, we per-formed two regression analyses using ConMax, RAI, and theirinteraction as independent variables and emotion and behaviorseparately as dependent measures. The overall equations weresignificant (for behavior, adjusted R2 = .51; for emotion, ad-justed R2 = .50, both ps < .0001). As expected, perceived con-trol, controlling for autonomy, contributed uniquely to bothbehavior (0 = .56, p < .001) and emotion {0 = .53, p < .001).More interestingly, autonomy also contributed uniquely to bothbehavior (0 = .25, p < .001) and emotion (0 = .27, p < .001),even controlling for perceived control. Neither interactionreached significance. Given the high zero-order correlationsbetween perceived control and the dependent variables, it wasnoteworthy that autonomy accounted for variance above andbeyond that of perceived control.

The next two sets of regressions examined the pattern ofunique relations between the subscales of each set of percep-tions and behavior and emotion. When the five componentsubscales of control (effort, ability, powerful others, luck, andunknown) were simultaneously regressed onto behavior, twoattributions accounted for unique variance in behavior: effort(0 = .49, p < .001) and powerful others (<8 = -.19, p < .001).When emotion was regressed on the five attributions, all butability contributed uniquely to the equation (effort, 0 = .28, p <.001; powerful others, 0= -.14, p< .02; luck, 0 = -.20, p <.005; and unknown, 0 = -.16, p < .005). The correspondinganalyses, in which the four components of autonomy were re-gressed on behavior and emotion, revealed that all the reasonsother than introjected uniquely predicted behavior (external, 0= - .21, p < .001; identified, 0 = .48, p < .001; intrinsic, 0= A3,p < .05); and emotion was uniquely predicted by all four sub-scales (external, 0 = -.15, p < .02; introjected, 0 = -.23, p <.001; identified, 0 = .35, p < .001; and intrinsic, 0 = .19, p <.005).

PERCEIVED CONTROL AND AUTONOMY 787

Table 4Correlations Between the Subcomponents of Perceived Control and Autonomy and BetweenBehavior and the Subcomponents of Emotion

Variable

Perceived controlConMaxEffortAbilityPowerful othersLuckUnknown

Behavior

RAI

.52

.53

.47-.40-.43-.30

Emotion.70

External

-.34-.34-.39

.27

.32

.30

Positive.63

Subcomponents

Autonomy

Introjected

-.17-.21-.32

.14

.15

.24

Emotion

Boredom-.64

Identified

.50

.49

.27-.32-.37-.21

Distress-.44

Intrinsic

.48

.47

.33-.36-.34-.15

Anger-.27

Note. N = 264 children, Grades 3-5. RAI = Relative Autonomy Index; ConMax = maximum controlprofile. Coefficients greater than .21 are significant at the .001 level; greater than .16 at the .01 level;greater than . 13 at the .05 level.

Next, to examine the unique contributions of all the compo-nents of both control and autonomy, we put the five attributionsand the four autonomy subscales together into one analysis andregressed behavior onto the full set of nine independent vari-ables (see Figure 1). As can be seen, the strongest unique predic-tors of behavior were effort attributions and identified reasons;powerful others and unknown also contributed uniquely to be-havior. When emotion was subsequently regressed onto thenine subscales, identified reasons emerged as the strongestunique predictor; other unique predictors of emotion includedeffort, powerful others, luck, and unknown attributions, andintrojected and intrinsic reasons (see Figure 1).

The number of unique predictors of emotion was striking

(everything except ability attributions and external reasons).However, the fact that the emotion composite included a well-differentiated set of subcomponents (positive emotions, bore-dom, distress, and anger) rather than a more general measure ofoverall emotional tone, may account for this finding. Specificattributions and reasons may relate to specific subcomponentsof emotion. To test this explanation, we ran a set of analyses inwhich each emotion subcomponent was considered separatelyas a dependent variable and regressed onto all nine subscales ofperceived control and autonomy simultaneously (Figure 2). Pos-itive emotion was uniquely predicted only by effort attributionsand by the more self-determined reasons, identified and intrin-sic. Anger was predicted by luck attributions and introjected

Table 5Correlations Between the Subcomponents of Perceived Control and Autonomyand the Subcomponents of Behavior and Emotion

Variable

Perceived controlConMaxEffortAbilityPowerful othersLuckUnknown

AutonomyRAIExternalIntrojectedIdentifiedIntrinsic

Behavior

.69

.70

.46-.55-.53-.45

.54-.32-.13

.57

.48

Emotion

.67

.61

.47-.51-.58-.50

.55-.37-.29

.47

.46

Positive

.61

.57

.38-.44-.47-.35

.50-.22-.08

.57

.56

Boredom

-.49-.47-.37

.39

.38

.39

-.50.38.22

-.36-.40

Distress

-.56-.49-.40

.43

.57

.49

-.37.35.40

-.21-.18

Anger

-.30-.25-.24

.21

.32

.30

-.23.19.32

-.10-.12

Note. N= 264 children, Grades 3-5. Coefficients greater than .21 are significant at the .001 level; greaterthan .16 at the .01 level; greater than .13 at the .05 level. RAI = Relative Autonomy Index; ConMax =maximum control profile.

788 B. PATRICK, E. SKINNER, AND J. CONNELL

Perceiued Control Butonomu

Figure I. Unique relationships between perceived control and auton-omy subcomponents (independent variables) and behavior and emo-tion composites (dependent variables). (Coefficients are standardizedregression coefficients. *p < .05. **p < .01. ***p < .001.)

reasons, whereas distress was a function of both luck and un-known attributions as well as of introjected reasons. Boredomwas predicted by unknown attributions and by both externaland lack of intrinsic reasons.

Motivational Profiles

Our specific predictions were that (a) children who were pas-sive and emotionally disaffected (low behavior and emotion)would be distinguished from children who were active andemotionally disaffected (low behavior and high emotion) on thebasis of their perceived control and (b) children who were activeand emotionally disaffected (high behavior and low emotion)would be distinguished from children who were active andemotionally engaged (high behavior and emotion) on the basisof their autonomy. To test these hypotheses, we constructed twonew dependent variables, one for each hypothesis, to reflectinteractions between behavior and emotion. First, we identi-fied students with extreme high or low values for behavior andemotion, using selection criteria that were designed to produceat least 20 children per group. First, we coded as 0 those stu-dents (« = 46) with low behavior (less than 3.0 on a 4-point scale,M= 2.55) and low emotion (less that 2.8 on a 4-point scale, M =2.35), and we coded as 1 those children (« = 23) with highbehavior (greater than 3.0 on a 4-point scale, M = 3.28) but lowemotion (less than 2.8 on a 4-point scale, M = 2.35).

We then regressed this new variable simultaneously onto per-ceived control and autonomy to determine whether they woulduniquely predict group membership. Perceived control distin-guished between the two groups (P = .30, p < .05), indicatingthat, controlling for autonomy, children with high behavior andlow emotion had higher perceived control than children withlow behavior and emotion. Autonomy, controlling for perceived

control, did not distinguish the groups. These results suggestthat, given a relatively unhappy student, increases in behaviorwill result from increases in control but not autonomy.

To test the second prediction, we contrasted two groups: thehigh behavior and low emotion group described above werecoded as 0; coded as 1 were students (n = 54) with high behavior(greater than 3.0 on a 4-point scale, M= 3.77) and high emotion(greater than 3.6 on a 4-point scale, M = 3.76). When we re-gressed this variable simultaneously onto ConMax and RAI,we found that both RAI and ConMax were strong unique pre-dictors (RAI p = .48; ConMax p = .40, both ps < .0001). Inother words, for students already high in behavioral engage-ment, both autonomy and perceived control predicted to betteremotional outcomes.

Discussion

Strong empirical support was found for the proposition thattwo distinct sources of motivation contribute uniquely to chil-dren's behavior and emotion during classroom learning activi-ties. Despite the powerful effects of perceived control on chil-dren's motivation, autonomy was able to account for additionalunique variance. Although the hypothesized interactions be-tween control and autonomy were not found in predicting be-havior and emotion, the pattern of regressions indicated thatthe effects of autonomy were not redundant with the effects ofcontrol, despite the large overlap between the two sets of per-ceptions. Instead, autonomy augmented the total amount ofvariance in motivation already accounted for by control.

Clear patterns emerged depicting the combinations of con-trol and autonomy that predict to behavior and specific emo-tions. Most behavioralfy active are children who believe in theefficacy of effort and whose reasons for task involvement are

Perceiued Control Emotion Hutonomu

Figure 2. Unique relationships between perceived control and auton-omy subcomponents (independent variables) and emotion subcom-ponents (dependent variables). (Coefficients are standardized regres-sion coefficients. *p < .05. **p < .01. ***p < .001. N = 264.)

PERCEIVED CONTROL AND AUTONOMY 789

identified. The expression of positive emotional engagement isa function of the two more autonomous reasons for engaging inlearning activities (identified and intrinsic) and, to a lesser ex-tent, of effort beliefs. Optimal motivation, then, characterizedby active behavioral involvement, interest, enthusiasm, and hap-piness, is the result of both perceived control centered on theeffectiveness of effort and reasons for engagement that are au-tonomous.

At the same time that both control and autonomy are uniquepredictors of children's feelings in the classroom, the preciseaspects of beliefs that contribute to specific emotions are distin-guishable. Although distress and anger form separate dimen-sions in the factor analysis, their predictors are similar. Bothworried and angry children's beliefs about control rely on luckeven though they see themselves as unlucky, and their reasonsfor task involvement are pressured internally (introjected); inaddition, they don't believe they know the strategies to producesuccess and avoid failure in school. Finally, children who arebored also report more external as well as fewer intrinsic rea-sons for engagement, and their control beliefs are organizedaround unknown strategies.

The link between external reasons and boredom makes clearthat lack of intrinsic motivation (also a predictor of boredom)does not imply the absence of reasons for task involvement.Instead, when intrinsic motivation is missing, children can mo-tivate themselves using several qualitatively different styles.And the one that leads to boredom are reasons that are exter-nal, that is, are based on expectations of rewards or fear ofpunishments from others. In addition, the links between intro-jection and both distress and anger point out the dual nature ofpressured, introjected reasons. On the one hand, the regulationof behavior is internalized, and so internal pressure to act incertain ways is maintained, resulting in distress. At the sametime, however, because regulation of behavior is not integrated,this style retains the evaluative character of being pressured byothers, resulting in anger.

Finally, both control and autonomy are predictors of motiva-tional profiles, or patterns of behavior and emotion. Of specialinterest are children who are behaviorally active but emotion-ally disaffected. These children differ from completely disaf-fected children (low behavior and low emotion) only in terms oftheir perceived control. In other words, children who feel thatthey have no control are passive, no matter what their reasonsfor engagement (Deci & Ryan, 1985). However, children whoare active but emotionally disaffected differ from fully engagedchildren (high behavior and emotion) on both autonomy andcontrol. For children who are already behaviorally active, morepositive emotions are the result of both higher perceived controland reasons for engagement that are more autonomous.

Limitations of the Present Study

The implications of these findings for theory and research onmotivation should be discussed in the context of the study'sstrengths and limitations. One strength is the measures of per-ceived control and autonomy. First, equally valid and reliableindicators were used, and so any differences in the relative pre-dictive power of the constructs cannot be due to differentially

sensitive instruments. Second, the assessments of perceivedcontrol and autonomy are differentiated, allowing conclusionsnot only about the general importance of the two sets of percep-tions, but also about the unique predictions from multipleaspects of each. It is interesting to note, for example, that theaspects of perceived control that predict to positive emotions(i.e., effort beliefs) are not the same ones that predict to distress(i.e., luck and unknown).

The major limitation of the study centers on the assessmentsof behavior and emotion. On the one hand, it would have beenhelpful to have an additional measure of student active behav-ioral engagement. Although both perceptions and emotions areinternal phenomena and most validly assessed through self-re-ports, children's behavioral engagement could also have beenreported by teachers and observers. Some evidence does indi-cate that student and teacher reporters converge in their assess-ments of behavior. For example, a study that compared teacher-and self-reports of student behavior for children of the sameages as in this study found them to be significantly correlated(Skinner & Belmont, in press). In addition, studies of the effectsof perceived control on teacher ratings of behavior and emotionreveal the same pattern of findings as the present study (Skin-ner et al., 1990). Nevertheless, it would be important to repli-cate the current pattern of relationships using other assess-ments of children's behavior.

The assessment of emotion, although reliable and valid, didnot tap as many dimensions of emotion as intended or certainlyas possible. Assessment of additional emotions might alsobroaden the picture; ones suggested by other theories wouldinclude pride, shame, guilt, humiliation, and embarrassment(Covington, 1984; Weiner, 1979, 1985). These emotions mayalso be more appropriate for students at slightly older ages thanthe third to fifth graders in the present study.

Implications for Intervention

The study leads to firm recommendations for remediation ofemotional disaffection in the classroom (Connell & Ryan, 1984;Ryan, Connell, & Deci, 1985). Most motivational interventionsfocus on structure, broadly defined, including increasing chil-dren's success experiences, increasing the contingency betweenchildren's behavior and important outcomes, and making surethat children have the strategies they need to produce successand avoid failure. Although these interventions may be effec-tive in increasing children's perceived control, they will, at best,be ineffectual in facilitating their autonomy. And at worst, theymay make children feel even more evaluated and pressured bythe relentless focus on success and outcomes. To facilitate au-tonomy, children need contextual support for autonomy (Con-nell & Ryan, 1984; Deci & Ryan, 1985). Autonomy support, adimension that is orthogonal to structure, includes provision ofchoice, lack of coercion, respect for children's own agendas, andlearning activities relevant to children's own goals. The currentconceptualization makes very clear that only in contexts thatsupport autonomy can children be fully engaged in learningtasks.

When it is not feasible to measure perceived control and au-tonomy directly, this research can also be used as a guide to"diagnosing" motivational problems directly from children's

790 B. PATRICK, E. SKINNER, AND J. CONNELL

patterns of behavior and emotion. For example, teachers cantest the idea that children who are behaviorally active but boredhave less intrinsic and more external reasons and so may needautonomy support in the form of more intrinsically interestingand relevant activities to facilitate more identified and intrinsicreasons. On the other hand, teachers might expect childrenwho are passive and anxious to be relying on luck and unknownstrategies, and so may provide more structure (perhaps in theform of more effective strategies) to improve motivation.

And of course, the current study suggests that any scheme forevaluating the effectiveness of a program to improve motiva-tion should not focus solely on behavior as is typically the case,but should also include the assessment of emotion and multipleaspects of control and autonomy. As has been seen, it would bepossible to improve behavioral engagement at the cost of auton-omy. In fact, many programs that seek to improve motivationthrough such strategies as competition or "raising the level ofconcern" (Hunter, 1982) are almost certainly making such asacrifice.

ConclusionsAn important implication of the current research is that

other sources of motivation, in addition to perceived control,influence children's behavioral and emotional engagement inthe classroom. Although no one would seriously argue that per-ceived control is the single predictor of motivation, the study ofother sources has not only been slowed by the empirical successof perceived control, but has also tended to be conducted inrelative isolation from research on perceived control.

The search for complementary or even synergistic predictorsof behavior and emotion may lead to broader, more integrative,theories of motivation. For example, one theoretical model sug-gests that three sources of motivation exist, based on three in-nate psychological needs (Connell, 1991; Connell & Wellborn,1991; Deci & Ryan, 1985). The three needs suggested by thismodel are competence, autonomy, and relatedness. This theorycould explain why perceived control and autonomy are indepen-dent sources of engagement and would also suggest that a thirdindependent source could be studied: the belief systems asso-ciated with people's needs to be connected with and loved byothers (Connell & Wellborn, 1991; Weiner, 1990, "belonging-ness"). Already, research conducted in the attachment traditionhas shown that internal working models of attachment figurespredict to exploratory behavior and emotional reactions tostress (Ainsworth, 1979). Whatever the source, however, itwould be important to examine the effectiveness of the newconstruct in predicting to both behavior and emotion, to exam-ine its effects in interaction with other sources of motivation,and to test its power in distinguishing among children withdifferent motivational profiles. Such a systematic approachmay be one route for increasing our understanding of the dy-namics of motivation.

References

Abramson, L. Y., Seligman, M. E. P., & Teasdale, J. D. (1978). Learnedhelplessness in humans: Critique and reformulation. Journal of Ab-normal Psychology, 87, 49-74.

Ainsworth, M. D. S. (1979). Infant-mother attachment. American Psy-chologist, 34, 932-937.

Bandura, A. (1986). The social foundations of thought and action: Asocial-cognitive theory. Englewood Cliffs, NJ: Prentice-Hall.

Chandler, C. L., & Connell, J. P. (1987). Children's intrinsic, extrinsic,and internalized motivation: A developmental study of behavioralregulation. British Journal of Developmental Psychology, 5,357-365.

Connell, J. R (1985). A new multidimensional measure of children'sperceptions of control. Child Development, 56, 1018-1041.

Connell, J. P. (1991). Context, self and action: A motivational analysisof self-system processes across the life-span. In D. Cicchetti (Ed.),The self in transition: From infancy to childhood. Chicago: Universityof Chicago Press.

Connell, J. R, & Ryan, R. M. (1984). A developmental theory of moti-vation in the classroom. Teacher Education Quarterly, II, 64-77.

Connell, J. P., & Ryan, R. M. (1987). Self-Regulatory Style Question-naire: A measure of external, introjected, identified, and intrinsic rea-sons for initiating behavior. Manual, University of Rochester.

Connell, J. P., & Wellborn, X G. (1991). Competence, autonomy, andrelatedness: A motivational analysis of self-system processes. InM. R. Gunnar & L. A. Sroufe (Eds.), Self processes in development:Minnesota Symposium on Child Psychology (Vol. 23, pp. 43-77).Chicago: University of Chicago Press.

Covington, M. (1984). The self-worth theory of achievement motiva-tion: Findings and implications. Elementary School Journal, 85,5-20.

DeCharms, R. (1976). Enhancing motivation: Change in the classroom.New \brk: Irvington.

DeCharms, R. (1981). Personal causation and locus of control: Twodifferent traditions and two uncorrelated constructs. In H. Lefcourt(Ed), Research with the locus of control construct. Vol. 1. Assessmentmethods (pp. 337-358). San Diego, CA: Academic Press.

Deci, E. L., & Ryan, R. M. (1985). Intrinsic motivation and self-determi-nation in human behavior. New York: Plenum Press.

Dweck, C. S., & Leggett (1988). A social-cognitive approach to motiva-tion and personality. Psychological Review, 95, 256-273.

Harter, S. (1981). A model of mastery motivation in children: Individ-ual differences and developmental change. In W A. Collins (Ed.),The Minnesota Symposium on Child Psychology (Vol. 14, pp. 215-255). Chicago: University of Chicago Press.

Hunter, M. (1982). Mastery teaching: Increasing instructional effective-ness in elementary, secondary schools, colleges and universities. ElSegundo, CA: TIP publications.

Nicholls, J. G. (1984). Achievement motivation: Conceptions of ability,subjective experience, task choice, and performance. PsychologicalReview, 91, 328-346.

Nolen-Hoeksma, S., Girgus, J., & Seligman, M. E. P. (1986). Learnedhelplessness in children: A longitudinal study of depression, achieve-ment, and explanatory style. Journal of Personality and Social Psy-chology, 51, 435-442.

Ryan, R. M. (1982). Control and information in the intrapersonalsphere: An extension of cognitive evaluation theory. Journal of Per-sonality and Social Psychology, 43, 450-461.

Ryan, R. M., & Connell, J. P. (1989). Perceived locus of causality andinternalization: Examining reasons for acting in two domains. Jour-nal of Personality and Social Psychology, 57, 749-761.

Ryan, R. M., Connell, J. P., & Deci, E. L. (1985). A motivational analy-sis of self-determination and self-regulation in education. In C.Ames & R. E. Ames (Eds.), Research on motivation in education: Theclassroom milieu. San Diego, CA: Academic Press.

Skinner, E. A. (1991). Development and perceived control: A dynamicmodel of action in context. In M. R. Gunnar & L. A. Sroufe (Eds.),Self processes in development: Minnesota Symposium on Child Psy-chology (Vol. 23, pp. 167-216). Chicago: University of ChicagoPress.

PERCEIVED CONTROL AND AUTONOMY 791

Skinner, E. A., & Belmont, M. J. (in press). Motivation in school: Recip-rocal effects of teacher behavior and student engagement across theschool year. Journal of Educational Psychology.

Skinner, E. A., Chapman, M., & Baltes, P. B. (1988). Beliefs aboutcontrol, means-ends, and agency: A new conceptualization and itsmeasurement during childhood. Journal of Personality and SocialPsychology, 54,117-133.

Skinner, E. A., & Connell, J. P. (1986). Development and the under-standing of control. In M. M. Baltes & P. B. Baltes (Eds.), Aging andthe psychology of control (pp. 35-63). Hillsdale, NJ: Earlbaum.

Skinner, E. A., Wellborn, J. G., & Connell, J. R (1990). What it takes todo well in school and whether I've got it: The role of perceived con-trol in children's engagement and school achievement. Journal ofEducational Psychology, 82, 22-32.

Weiner, B. (1979). A theory of motivation for some classroom experi-ences. Journal of Educational Psychology, 71, 3-25.

Weiner, B. (1985). An attributional theory of achievement and emo-tion. Psychological Bulletin, 92, 548-573.

Weiner, B. (1990). History of motivational research in education. Jour-nal of Educational Psychology, 82, 616-622.

Wellborn, J. G. (1991). Engaged and disaffected action: The conceptual-ization and measurement of motivation in the academic domain. Un-published doctoral dissertation, University of Rochester, RochesterNY.

Wellborn, J. G., & Connell, J. P. (1987). Manual for the Rochester As-sessment Package for Schools. University of Rochester, Rochester,NY.

Wellborn, J. G., Connell, J. P., & Skinner, E. A. (1989). The StudentPerceptions of Control Questionnaire (SPOCQ). Technical Report,University of Rochester, Rochester, NY.

White, R. W (1959). Motivation reconsidered: The concept of compe-tence. Psychological Review, 66, 297-333.

Received November 11,1992Revision received March 15,1993

Accepted March 16,1993 •