20
School Psychology Review, 2004, Volume 33, No. 3, pp. 344-362 Extrinsic Reinforcement in the Classroom: Bribery or Best Practice K. Angeleque Akin-Little University of the Pacific Tanya L. Eckert and Benjamin J. Lovett Syracuse University Steven G. Little University of the Pacific Abstract. The debate over the effects of the use of extrinsic reinforcement in class- rooms, businesses, and societal settings has been occurring for over 30 years. Some theorists have cautioned against the use of reward, whereas others have found little, if any, detrimental effect. This article examines the debate with an emphasis on data-based findings. The extrinsic/intrinsic dichotomy is explored along with seminal studies in both the cognitive and behavioral literatures. The results from important meta-analytic studies are presented. From this review, it is concluded that little detrimental effect is found with the use of external reinforcement. Read- ers are given specific recommendations on the appropriate use of reinforcement programs in educational settings. Many educational personnel have at least some rudimentary knowledge of the effects of rewards and/or reinforcement on students' be- havior in school settings. Observations of class- rooms and school settings frequently reveal evidence of some sort of reward system for academic output and/or appropriate behavior. For example, stickers may be given to students for completed assignments or pizza coupons may be given for appropriate classroom behavior. Schools have successfully employed the use of external rewards for decades (Slavin, 1997). The past 40 years have witnessed the success of the use of reinforcement procedures in the classroom (Allyon & Azrin, 1968; Barrish, Saunders, & Wolf, 1969; Buisson, Murdock, Reynolds, & Cronin, 1995; Cavalier, Ferretti, & Hodges, 1997; O'Leary & Drabman, 1971; Swiezy, Matson, & Box, 1992). Along with the research on the effective- ness of external reinforcers in the schools, there has been a rise in concern on the part of some educators and psychologists over the use of reward contingency systems in classrooms across the country (Deci, Koestner, & Ryan, 1999a, 1999b, 2001; Kohn, 1993, 1996). The problem, these researchers assert, is the effect an extrinsic reinforcer may have on a student's intrinsic motivation to perform a reinforced task once the reinforcer for that task is with- drawn. These researchers speculate that if re- inforcement strategies are used, an individual's perceptions of competence and self-determi- nation will decrease, thereby decreasing that Correspondence regarding this article should be addressed to Angeleque Akin-Little, PhD, 101 Benerd School of Education, University of the Pacific, 3601 Pacific Avenue, Stockton, CA 95211; E-mail: [email protected] Copyright 2004 by the National Association of School Psychologists, ISSN 0279-6015 344

Extrinsic Reinforcement in the Classroom: Bribery or …misdmtssnetwork.weebly.com/.../1/0/5/9/1059434/extrinsic_rewards.pdf · Extrinsic Reinforcement in the Classroom: ... evidence

Embed Size (px)

Citation preview

School Psychology Review,2004, Volume 33, No. 3, pp. 344-362

Extrinsic Reinforcement in the Classroom: Briberyor Best Practice

K. Angeleque Akin-LittleUniversity of the Pacific

Tanya L. Eckert and Benjamin J. LovettSyracuse University

Steven G. LittleUniversity of the Pacific

Abstract. The debate over the effects of the use of extrinsic reinforcement in class-rooms, businesses, and societal settings has been occurring for over 30 years. Sometheorists have cautioned against the use of reward, whereas others have foundlittle, if any, detrimental effect. This article examines the debate with an emphasison data-based findings. The extrinsic/intrinsic dichotomy is explored along withseminal studies in both the cognitive and behavioral literatures. The results fromimportant meta-analytic studies are presented. From this review, it is concludedthat little detrimental effect is found with the use of external reinforcement. Read-ers are given specific recommendations on the appropriate use of reinforcementprograms in educational settings.

Many educational personnel have at leastsome rudimentary knowledge of the effects ofrewards and/or reinforcement on students' be-havior in school settings. Observations of class-rooms and school settings frequently revealevidence of some sort of reward system foracademic output and/or appropriate behavior.For example, stickers may be given to studentsfor completed assignments or pizza coupons maybe given for appropriate classroom behavior.Schools have successfully employed the use ofexternal rewards for decades (Slavin, 1997). Thepast 40 years have witnessed the success ofthe use of reinforcement procedures in theclassroom (Allyon & Azrin, 1968; Barrish,Saunders, & Wolf, 1969; Buisson, Murdock,Reynolds, & Cronin, 1995; Cavalier, Ferretti,

& Hodges, 1997; O'Leary & Drabman, 1971;Swiezy, Matson, & Box, 1992).

Along with the research on the effective-ness of external reinforcers in the schools, therehas been a rise in concern on the part of someeducators and psychologists over the use ofreward contingency systems in classroomsacross the country (Deci, Koestner, & Ryan,1999a, 1999b, 2001; Kohn, 1993, 1996). Theproblem, these researchers assert, is the effectan extrinsic reinforcer may have on a student'sintrinsic motivation to perform a reinforcedtask once the reinforcer for that task is with-drawn. These researchers speculate that if re-inforcement strategies are used, an individual'sperceptions of competence and self-determi-nation will decrease, thereby decreasing that

Correspondence regarding this article should be addressed to Angeleque Akin-Little, PhD, 101 BenerdSchool of Education, University of the Pacific, 3601 Pacific Avenue, Stockton, CA 95211; E-mail:[email protected]

Copyright 2004 by the National Association of School Psychologists, ISSN 0279-6015

344

Extrinsic Reinforcement in the Classroom

individual's intrinsic motivation to perform thetask. For example, in some teacher guidebooks,teachers are told that the use of extrinsic rein-forcement can decrease creativity (Tegano,Moran, & Sawyers, 1991). Further, manyteacher education programs embrace a cogni-tive theory of education (e.g., Bruner, 1960)that emphasizes intuition and insight to facili-tate learning. In the resulting teaching prac-tices (e.g., discovery leaming, constructivism),the teacher does not impart knowledge; rather,the focus is on arranging the environment tohelp students "discover" knowledge. The ac-cent is on internal, intrinsic machinations withno external reinforcement procedures used.This pedagological instruction may be in di-rect conflict with research supporting the useof external reinforcers in the classroom and theefficacy of direct instruction (Alberto &Troutman, 2003). Finally, Kohn (1993) goesas far as to state that the use of external re-wards, even verbal praise, can be consideredbribery to invoke temporary obedience andmake children dependent on adult approval.This perspective is prevalent not only in teachereducation programs, but in society as a whole.

The debate regarding the use of extrin-sic reinforcers began in the 1970s with studiesattempting to examine the effects of reward onan individual's intrinsic motivation. The debategained new impetus in 1994 when Cameron andPierce conducted a meta-analysis on this topic(Cameron & Pierce, 1996; Kohn, 1996; Lepper,1998; Lepper, Keavney, & Drake, 1996; Ryan& Deci, 1996). More recently, two additionalmeta-analytic studies have been conducted toexamine the effects of extrinsic reinforcers onintrinsic motivation (Cameron, Banko, &Pierce, 2001; Deci et al., 1999a). The resultshave been contradictory with some reviewersfinding no detrimental effect or detrimentaleffect only under certain prescribed conditions(Camieron & Pierce, 1994; Cameron et al.,2001) and other researchers finding negativeeffect (Deci et al.). Further, other reviewershave attempted to provide illumination for con-tradictory findings by examining findings ofdetrimental effect from a behavioral, scientificperspective (Akin-Little & Little, 2004; Car-ton, 1996; Dickinson, 1989; Flora, 1990).

The present article brings this debate intothe school psychology literature and attemptsto review the evidence of the effectiveness ofreinforcement programs as currently used inschool settings. An attempt is made to definethe terms intrinsic and extrinsic motivationfrom a historical perspective. Cognitive inves-tigations that form the basis for criticisms ofthe use of reward are discussed and critiquedin terms of methodological concerns, the useof the construct "intrinsic motivation," andwhether or not it is prudent to study an unob-servable construct. Alternative, behavioralexplanations for perceived decrements are alsopresented. Findings from behavioral investi-gations (i.e., studies using repeated measures)that have tended to be ignored are detailed andrecent meta-analytic findings (i.e., Cameron etal., 2001; Deci et al., 1999a) are discussed. Bestpractice suggestions for both teachers andschool psychologists are offered on the use ofextrinsic reinforcement in the classroom. Fi-nally, ideas for future research are provided.

Defining Intrinsic and ExtrinsicMotivation

An intrinsically motivated behavior hasbeen defined by Deci (1975) as one for whichthere exists no recognizable reward except theactivity itself (e.g., reading). That is, behaviorthat cannot be attributed to external controls isusually attributed to intrinsic motivation. Con-sequently, an extrinsically motivated behaviorrefers to behavior controlled by stimuli externalto the task. The work of Harlow, Harlow, andMeyer (1950) is generally cited as the first at-tempt to distinguish between intrinsic motiva-tion and externally rewarded behavior. Harlowet al. studied rhesus monkeys given a puzzle-solving task. Because the monkeys solved thetask with no identifiable reward (e.g., food, wa-ter), the researchers extrapolated from this thatthe monkeys' intrinsic motivation must accountfor their exploratory and manipulative behav-ior. Based upon these findings, behavior exhib-ited when using external rewards (i.e., raisins)was labeled as behavior extrinsically motivatedand behavior exhibited when no rewards wereobserved was labeled as behavior intrinsicallymotivated. Subsequent researchers have also at-

345

School Psychology Review, 2004, Volume 33, No. 3

tempted to define intrinsically motivated be-havior. These definitions have included defm-ing intrinsic motivation as the need forachievement (McClelland, Atkinson, Clark,& Lowell, 1953), the need to be effectiveand competent in dealing with one's envi-ronment (White, 1959), the need to conceiveof oneself as the locus of causality(DeCharms, 1968), or the need to be self-deter-mining and competent (Deci).

Not surprisingly, a debate has resultedsurrounding the distinction between intrinsicand extrinsic motivation. Several critics (e.g.,Guzzo, 1979; Scott, 1975) have produced datathat illuminate the problems associated withidentifying intrinsically motivated behaviors.Other theories have been proposed that pur-port to explain behavior that appears to occurin the absence of any extrinsic motivation.These behaviors may, in fact, be due to antici-pated future benefits (Bandura, 1977) or inter-mittent reinforcement (Dickinson, 1989).Some have suggested that cognitive definitionsof intrinsic motivation are definitions "by de-fault" (Zimmerman, 1985). In other words,when no extemal reward mechanism is foundcontrolling a particular behavior, the motiva-tion for that behavior is identified as intrinsi-cally controlled. Additionally, the criticism thatcontingent consequences may decrease intrin-sic motivation is based on the belief that a"sharp distinction can be made between be-havior maintained by obvious environmentalconsequences-extrinsically motivated behav-ior-and behavior that occurs in the seemingabsence of consequences-intrinsically moti-vated behavior" (Dickinson, 1989, p. 1).

According to Deci's (1975) definition,intrinsic motivation is evidenced when peopleparticipate in an activity because of the inter-nal enjoyment of the activity and not be-cause of any perceived extrinsic reward.That is, behavior that cannot be attributedto external controls is usually attributed tointrinsic motivation. Intrinsic motivationenables people to feel competent and self-determining. Intrinsically motivated behav-ior is said to result in creativity, flexibility,and spontaneity. In contrast, extrinsically mo-tivated actions are characterized by pressure

and tension, and are believed to result in lowself-esteem and anxiety (Deci & Ryan, 1985).

Distinctions between extrinsic and intrin-sic consequences can also be found in writ-ings by Horcones (1987) and Mawhinney,Dickinson, and Taylor (1989). Horcones statesthat intrinsic consequences occur in the absenceof programming by others. They are naturaland automatic responses inevitably producedby the structural characteristics of the physi-cal environment in which humans exist. Ex-trinsic consequences, conversely, are those thatoccur in addition to any intrinsic consequencesand are most often programmed by others (i.e.,the social environment, researchers, teacher,applied behavior analysts). Consequently,Mawhinney et al. (1989) offer the followingdefinition: "Intrinsically controlled behaviorconsists of behavior controlled by unpro-granuned consequences while extrinsically con-trolled behavior consists of behavior controlledby programmed consequences" (p. 111). Al-though early theorists and experimenters(Harlow et al., 1950) have attempted to explainbehavior through the use of the unobservableconstruct, a definition of intrinsic motivationthat all can agree upon has yet to be written.Clearly, this is related to competing theoreti-cal orientations (e.g., behavioral, which calls forobservable phenomena and social cognitive,which accepts nonobservable phenomena).

In general, if the dichotomy between in-trinsic and extrinsic motivation is accepted,intrinsic motivation is assumed to be of greatervalue (Fair & Silvestri, 1992). This belief isdue in large part to the Westernconceptualization of the human as autonomousand individualistic. In this view, humans aredriven toward self-actualization and any oc-currence that impinges on self-determninationcauses dissonance. Further, the use of extrin-sic reinforcement is seen as controlling and/orlimiting self-discovery, creativity, and the ca-pacity for humans to reach fulfillment(Eisenberger, Pierce, & Cameron., 1999). In-terestingly, when this tenet is examined in re-lationship to the use of punishment, punishmentis perceived as less of a threat to autonomy be-cause humans may choose how to behave toavoid punishment (Maag, 1996. 2001).

346

Extrinsic Reinforcement in the Classroom

Cognitive Investigations of Intrinsic/Extrinsic Motivation

Deci's Initial Studies

In three different experiments conductedby Deci in 1971, the "effects of externally medi-ated rewards on intrinsic motivation" (p. 105)were explored. These studies sought to answerthe question: Will intrinsic motivation decrease,increase, or remain the same after extrinsic re-wards have been experienced? All three experi-ments were based on the same general paradigmwith behavioral observations taking place dur-ing three different sessions. The participants firstperformed the tasks without any external reward.Next, half of the participants were rewarded forperforming the activity, and half of the partici-pants received no reward. Finally, rewards werewithdrawn and further observations were made.

In the first experiment, college studentsparticipated in three separate sessions con-ducted over a3-day period. During each 1-hoursession, the participants were presented with apuzzle-solving task that appeared to be intrin-sically interesting to college students. The firstsession consisted of all participants perform-ing the task without any offer of reward. Dur-ing the second session, experimental partici-pants were promised $1.00 for each puzilesolved. The third session was identical to the firstsession, and the experimental participants weretold that a reward was unavailable. Intrinsicmotivation was defined as the number of sec-onds participants spent on the task during free-choice periods. These periods occurred when theexperimenter left the room in the middle of asession and informed the participants that theycould choose to work on their activity. Resultsof this study indicated that the presentation ofrewards increased the experimental participants'time on task (i.e., extrinsic motivation),whereas the withdrawal of rewards decreasedthe experimental participants' time on task (i.e.,intrinsic motivation). During the final phaseof the study, significant differences in time ontask were observed between the experimentaland control participants, with the control par-ticipants engaging in more time on task.

In the second experiment, Deci (1971)examined the amount of time college student

participants spent completing newspaper head-lines. Similar to the findings of the previouslyreviewed study, Deci reported that the induce-ment of rewards increased the experimentalparticipants' extrinsic motivation for the ac-tivity, but the withdrawal of rewards under-mined the experimental participants' intrinsicmotivation for the task. Deci concluded thatthe results from both of these experiments sup-ported the hypothesis that external rewardsnegatively effect intrinsic motivation. He sug-gested that intrinsic motivation will decreaseif an extrinsic reward is obtained in a situationwhere individuals normally perceive them-selves to be the origin of their behavior.

In the third experiment, Deci (1971) rep-licated many of the procedures reported in thefirst experiment; however, verbal praise (i.e.,very good, much better than average for thisconfiguration) was substituted for tangible re-wards. Results of this experiment were similarto those reported in the first experiment. Partici-pants assigned to the experimental group spentmore time on task than participants assigned tothe control condition. Deci also measuredwhether or not the participants found the task tobe enjoyable and interesting. Using a 9-point self-rating scale, Deci found scores virtually identi-cal to the findings of the first experiment; thatis, participants seemed to find the task enjoy-able and interesting. Based on data collectedfrom Experiment 3, Deci found that the experi-mental group spent what he labeled significantlymore time on the task (difference scores betweenSession 3 and Session 1) than those who receivedno praise. According to Deci, these results sug-gested that social rewards do not decrease anindividual's intrinsic motivation to perform anactivity, and may even enhance intrinsic inter-est. From the results of these experiments, cog-nitive evaluation theory (Deci & Ryan, 1985)was developed.

Cognitive Evaluation Theory

Deci and Ryan's (1985) cognitive evalu-ation theory is based on the assumption thatself-determination and competence are innatehuman needs. Cognitive evaluation theorystates that events facilitate or hinder feelingsof competence and self-determination depend-

347

School Psychology Review, 2004, Volume 33, No. 3

ing on their perceived informational, controlling,or amotivational significance. Deci and Ryandivide rewards into two categories: task-contin-gent rewards that are rewards given for partici-pation in an activity, solving a problem, or com-pleting a task; and quality-dependent rewardsthat involve the "quality of one's performancerelative to some normative information or stan-dard" (p. 74). Task-contingent rewards are hy-pothesized to detrimentally affect intrinsic mo-tivation by decreasing self-determination (i.e.,reward is viewed as a controlling event attempt-ing to determiine behavior thereby decreasingself-determination and, consequently, intrinsicmotivation). Quality-dependent rewards are alsobelieved to act to decrease intrinsic motivationby reducing one's feelings of self-determination.However, quality-dependentrewards also serveto increase feelings of competence according toDeci and Ryan (i.e., reward is viewed as an in-formational event indicating skill at a certain task,leading to an increase in feelings of competence,which serves to increase intrinsic motivation).Therefore, it is never clear whether the detri-mental effect to self-determination or the in-cremental effect to competence will be stron-ger in experiments examining quality-depen-dent rewards. Thus, for Deci and Ryan, qual-ity-dependent rewards may not decrease in-trinsic motivation. The detrimental effect ofgreatest concem is in circumstances involvingtask-completion rewards.

Eisenberger and Cameron (1996) furtherdivide the task-completion rewards category intothe subcategories of performance-independentrewards that individuals receive simply for par-ticipation in an activity, and completion-inde-pendent rewards given when an individual hasfinished a task or activity. Cognitive evaluationtheory suggests that an individual's intrinsicmotivation would be most detrimentally affectedupon receipt of tangible, anticipated rewards.Additionally, according to this theory, verbalrewards may be informational, and therefore,increase intrinsic motivation as evidenced byDeci's (1971) findings in Experiment 3. Eventsmay also be perceived as amotivational indicat-ing an individual's lack of skill, which reducesone's cognitions of competence and, subse-quently, intrinsic motivation.

In 1988, Rummel and Feinberg con-ducted a meta-analysis assessing cognitive evalu-ation theory. They concluded that controlling,extrinsic rewards do have a damaging effect onintrinsic motivation, thus providing support forthe theory. Basic problems with cognitive evalu-ation theory, however, were also identified. First,faulty reasoning was used because rewards wereidentified as either controlling, informational, oramotivational after performance had been mea-sured. Second, feelings of competence and self-determination, central to the theory as agents forchange in intrinsic motivation, are not measur-able. The assumption was made that changeswere occurring because changes in behaviorwere observed. The constructs of self-determi-nation, competence, and even intrinsic motiva-tion were inferred from the very behavior theysupposedly cause (Cameron & Pierce, 1994).Additionally, the theory contains no explanationfor why the anxiety associated with a decreasein self-determination would reduce intrinsicmotivation. As Eisenberger and Cameron (1996)write, "based on the theory's premise, one couldaltematively argue that reduced self-determina-tion would, for example, reduce preference forthe reward or instigate anger at the person deliv-ering the reward" (p. 1156).

Results of a meta-analysis performed byCameron and Pierce (1994) partly serve to re-fute cognitive evaluation theory. Deci and Ryan(1985) stress the importance of measurementsof attitude because they theorize that interest,enjoyment, and satisfaction are central emotionsto intrinsic motivation. How a personfeels aboutan activity is reflected behaviorally as time spenton task. The results of the Cameron and Piercemeta-analysis, however, suggest that reward (andsubsequent withdrawal) tends not to affect atti-tude. These researchers further found that atti-tude seems to be affected positively when ver-bal rewards are used, and when rewards arecontingent on a precise level of achievement.

Researchers (Cameron & Pierce, 1994;Eisenberger & Cameron, 1996; Eisenberger etal., 1999) have suggested that cognitive evalu-ation theory is not a useful or viable theoryand that any decrements in behavior are betterexplained through leamed helplessness or gen-eral interest theory. In leamed helplessness, thedecrement in intrinsic motivation is said to be

348

Extrinsic Reinforcement in the Classroom

due to the single reward delivery paradigm uti-lized by most studies in this area. General inter-est theory suggests that intrinsic motivation isdriven by more thanjust self-determiination andcompetence needs. Eisenberger et al. proposethat rewards must be examined for both contentand context of tasks. Rewards that communi-cate task performance and satisfy needs, wants,and desires can increase intrinsic motivation,whereas rewards that convey a message that thetask is extraneous to needs, wants, and desiresmay decrease intrinsic motivation.

The Overjustification Hypothesis

Another experiment designed and con-ducted to explore the detrimental effects of rein-forcement on intrinsic motivation was the workof Lepper, Greene, and Nisbett (1973). Theseresearchers, using Bem's (1967) attributionalmodel, examined individuals currently engag-ing in a task or activity without the possibility ofextemal rewards. They hypothesized that introduc-tion of extrinsic rewards that can be earned for en-gaging in that task or activity (i.e., overly sufficientextrinsic pressure) may lead the individual to viewhis or her actions as extrinsically motivated. Con-sequendy, these individuals may find the activity,in the later absence of these extrinsic rewards, tobe of less intrinsic interest (Lepper, 1983).

In this experiment (Lepper et al., 1973),preschool children were chosen based on theirhigh baseline level of interest in drawing. Theparticipants were divided into three groups: anexpected-reward group, an unexpected-rewardgroup, and a no-reward group. Children in thefirst group were promised and received a good-player award contingent upon their drawingwith magic markers. Children in the secondgroup received an award, but were not prom-ised it beforehand, and children in the thirdgroup did not expect or receive an award.

In subsequent free-play sessions, chil-dren from the expected-reward group were ob-served to spend considerably less time drawingthan children from the other two groups. Theunexpected-reward and no-reward groupsshowed slight increases in drawing time. Lepperet al. (1973) concluded that their results providedevidence of an undesirable consequence of theuse of extrinsic rewards. However, this con-

clusion does not appear to be supported by thedata. If the receipt of the reward were the causeof a decrease in drawing behavior, one wouldexpect both the expected- and unexpected-re-ward group to exhibit a decrement in drawingbehavior. This was not the case.

In an attempt to explain their results,Lepper et al. (1973) offered the overjustificationhypothesis. According to this hypothesis, if aperson is already performing an activity and re-ceiving no extrinsic reward for that performance,introduction of an extrinsic reward will decreaseintrinsic interest or motivation. This occurs be-cause the person's performance is nowoverjustified, resulting in-the person's percep-tion that his or her level of intrinsic motivationto perforn the activity is less than it was ini-tially. According to this theory, the person sub-sequently performs the activity less once the re-inforcement is removed (Wllliams, 1980).

Lepper et al.'s (1973) findings have beensupported (e.g., Deci & Ryan, 1985; Greene &Lepper, 1974; Morgan, 1984) except with aca-demic tasks and older students. This suggeststhat an undermining effect of reward does notoccur if the students are told they have achieveda preset standard and the task is at a challenginglevel for them (Pittman, Boggiano, & Ruble,1983). Rewards have also been shown to exhibitan additive effect on intrinsic motivation whengiven dependent upon behavior (e.g., Lepper,1983), rewards provided information about thestudents' competence (e.g., Lepper & Gilovich,1981; Rosenfield, Folger, & Adelman, 1980),and rewards were given to students not optimalymotivated toward desirable educational goals(Morgan, 1984). Moreover, researchers haveconsistently found that verbal rewards tend toincrease intrinsic motivation, whereas tangiblerewards may decrease intrinsic motivation(Cameron & Pierce, 1994). Additionally, reduc-tions of intrinsic motivation have not been foundwith traditionally behavioral studies utilizing arepeated measures design (Cameron & Pierce).

Criticisms of Cognitive Research

Methodological Concerns '

The findings from the early studies ex-amining the use of external reinforcers have

349

School Psychology Review, 2004, Volume 33, No. 3

had a profound effect on education and society.It is important to carefully and objectively ex-amine the methodology. Harlow et al. (1950)concluded that the introduction of an externalreinforcer decreased intrinsic motivation. How-ever, the observation period associated with thereward phase in this study was only 5 minutesin duration. As a result, the observed effects mayhave been limited. Further, and more importantly,Harlow et al.'s subsequent research (see Davis,Settlage, & Harlow, 1950) found differing re-sults. That is, an overall increase in puzzle-solv-ing behavior was observed following the intro-duction of food and no decrease in puzzle-solv-ing behavior was observed (Zimmerman, 1985).

Cognitive research has also been criti-cized for failing to make any distinction be-tween rewards and reinforcement. These twoterms cannot be used synonymously. A rein-forcer is an event that increases the frequencyof the target behavior it follows, and a rewardis a pleasant occurrence that has not beenshown to necessarily strengthen behavior(Cameron & Pierce, 1994). The experimentsdetailed above (i.e., Deci, 1971; Lepper et al.,1973) utilized only one administration of re-ward and failed to document the reward usedwas indeed functioning as a true reinforcer.

Additionally, Feingold and Mahoney(1975) attempted to assess the validity of thecognitive theory of overjustification. They ex-amined the Deci (1971) and Lepper et al.(1973) studies and offered several criticismsof the cognitive research including seriousdoubts about internal and external validity, anda neglect of current literature on reinforcementcontrast. Feingold and Mahoney stated that theDeci and Lepper et al. experiments were notsimilar to the token economies found in mostclassrooms at that time. For example, the ex-perimental phases were extremely brief (i.e., asingle session), and the reinforcement effectwas not reported. These concerns raised ques-tions regarding the external validity of the find-ings. In addition, the internal validity of thestudy was criticized due to the observationtechniques employed (i.e., nonindependentobservers) as well as the marginally signifi-cant results reported (p < .10). Furthermore,Feingold and Mahoney reported that many of

the cognitive experiments failed to take intoconsideration the literature on reinforcementcontrast that accounts for previous response-consequence experiences. That is, if a behav-ior is weakly reinforced, then strongly rein-forced, and finally, reinforcement is returnedto the original weak state, performance sup-pression may occur. Feingold and Mahoneyconcluded that "a formerly reinforcing stimu-lus can become a punisher through relativecontrast" (p. 369), a contention virtually ig-nored by many of the cognitive experimentsof Deci and Lepper and colleagues. Finally, manyof the cognitive experiments did not take intoconsideration the role that anticipation of thereward may have played in their findings. Thisis especially true of the Lepper et al. experimentwhere one group of children was told they wouldreceive a reward and another group was not.Children in the unexpected-reward groups, whowould have been expected to show a decreasein interest, actually exhibited a slight increase inpre- to postperformance indicating an increasein intrinsic motivation, and no overjustificationeffect (Feingold & Mahoney).

Many authors (e.g., Dickinson, 1989)have criticized continued attempts to identifythe construct of "intrinsic motivation," suggest-ing that such efforts impede the goal of thescientific study of behavior. According to Flora(1990), no behavior occurs without an identi-fying external circumstance:

A complete scientific explanation of behav-ior does not require reference to constructswhich are, in principle, unobservable... Acomplete scientific account for any behav-ior of any organism may be obtained with acomplete description of the functional inter-dependency of the behavior-environmentinteraction. (p. 323)

By creating internal constructs that dependupon inferences in their behavioral explana-tions, some may indeed be obstructing the dis-covery of the true function of behavior throughmore scientific, measurable, observable means.

Alternative Explanations

A neglect of behavioral principles thatmay account for any decrements in observedbehavior has been common in the majority ofpast studies on intrinsic motivation (McGinnis,

Extrinsic Reinforcement in the Classroom

1996). Alternative explanations, more behav-ioral in nature, include anticipated future ben-efits (Bandura, 1977), intermittent reinforce-ment (Dickinson, 1989), competing responsetheory (Reiss & Sushinsky, 1975), behavioralcontrast (Bates, 1979; Feingold & Mahoney,1975), and discriminative stimuli (Flora, 1990).

Reiss and Sushinsky (1975) are espe-cially critical of the overjustification hypoth-esis, stating that the theory is too vague to beuseful for scientific purposes and that compet-ing response theory more adequately accountsfor any obtained decrements in intrinsic moti-vation. Competing response theory suggests thata student's intrinsic motivation may decreasebecause of other stimuli present in the environ-ment. Students respond to these stimuli, and thisresults in a decrease of theirresponse to the tar-geted activity before termination of contingen-cies occurs. Bates (1979) offers behavioral con-trast as an additional explanation for decrementsin intrinsic motivation. In this paradigm, twobehaviors are reinforced on different schedules.One behavior is then extinguished. This producesan increase in response of the other behavior.The classic example is of the pigeons pecking atdifferent colors. When the reinforcer for peck-ing at one color is withheld, the pecking at theremaining color increases in rate and intensity.Finally, Flora (1990) discusses the possibility ofdiscriminative stimuli as another explanation.According to this account, behaviors occur inan environmental context. Instead of examiningan unobservable construct such as intrinsic mo-tivation, Flora suggests it is more useful to de-termine the discriminative stimulus and the re-inforcers in the environment that maintain a func-tional relationship. These factors, Flora proposes,maintain a behavior's rate and occurrence.

Additionally; Dickinson (1989) proposedthat decrements in intrinsic motivation mayoc,cur if the activity is one that participants findboring or uninteresting, rewards are given foractivities culturally praised as intrinsicallymotivated behaviors (e.g., artistic or creativeactivities), or rewards become aversive stimuli.In the first instance, motivation is decreasedbecause satiation is reached through repeatedexposure to sensory reinforcement. In the sec-ond illustration, decrement is explained

through an examination of cultural norms.People are often praised if they engage in cer-tain activities that supposedly offer specificintrinsic rewards (e.g., painting, dancing). Ifan individual is then extrinsically rewarded forthis activity, the person may experience a de-crease in praise. If praise is reinforcing for thatperson, he or she may engage in the activityless because the activity is now differentiallycorrelated with the loss of praise. In the thirdexample, the subject may not participate in theactivity because they are angry with the ex-perimenter for withholding the reward, they failto meet the performance standards, or individu-als are offered rewards for engaging innonpreferred activities, and/or threatened withpunishment for noncompliance (Dickinson).

Eisenberger and Cameron (1996) alsopresent an interpretation of the specified con-ditions under which rewards may decrease in-trinsic motivation. They state that individualswho receive performance-independent rewardsmay perceive that they have no control overthe reward. This perception may lead to a de-crease in performance that may be misinter-preted as a decrease in intrinsic motivation.These authors suggest that the intrinsic inter-est decrement may be better explained bylearned helplessness that asserts that "uncon-trollable aversive stimulation results in gener-alized motivational deficits" (p. 1156). Learnedhelplessness theory predicts a decrease in in-trinsic motivation for performance-indepen-dent rewards. However, unlike cognitive evalu-ation theory, no prediction of a decrement issuggested following task-completion rewards.

Carton (1996) examined the cognitivistassertion that praise appears to increase intrin-sic motivation and the delivery of tangible re-wards appears to decrease intrinsic motivation.Again, these assumptions are based upon cog-nitive evaluation theory (Deci & Ryan, 1985).However, operant psychologists' reviews of theliterature on the effects of rewards on intrinsicmotivation (e.g., Bernstein, 1990; Dickinson,1989; Flora, 1990; Scott, 1975) reach vastly dif-ferent conclusions than those conducted by psy-chologists with decidedly cognitive viewpoints.Important points raised by operant psychologistsinclude the finding that many cognitivists have

351

School Psychology Review, 2004, Volume 33, No. 3

presumed that the use of reinforcers decreasesintrinsic motivation when in fact the rewardsutilized in these particular studies often did notshow a clear increase in response rate. Thus, bydefinition, the use of these presumed rewardswas not reinforcement. Furthermore, cognitivestudies did not assess response rates for stabil-ity, behavioral observations included in most ofthese studies were oftenrelatively brief, and thesestudies rarely included follow-up observations.Carton's review of the literature finds little sup-port for examples of a decrease in intrinsic mo-tivation based on the cognitive evaluation theoryand reveals three confounding effects: (a) tem-poral contiguity, (b) the number of reward ad-ministrations, and (c) discriminative stimuli as-sociated with reward availability.

The effects of the number of rewardadministrations (i.e., repeated delivery of re-wards is more likely to produce an increase in atarget behavior as opposed to a single adminis-tration) and discriminative stimuli associatedwith reward availability (Flora, 1990) have beenpreviously discussed in this article. However,Carton's (1996) notion of the effects of tempo-ral contiguity has not. Temporal contiguity re-fers to the amount of time between the occur-rence of the target behavior and the delivery ofthe consequence. In an examination of the lit-erature, Carton found time differences betweenthe delivery of tangible rewards and verbal re-wards (i.e., praise) in many studies. Most of theverbal rewards were delivered immediately af-ter the target behavior occurred, thereby increas-ing the likelihood that behavior would be re-peated. In contrast, tangible rewards were oftendelivered days or weeks after treatment, virtu-ally ensuring a decrease in the occurrence of thetarget behavior. Carton's finding that research-ers in those studies have consistently found de-creases in intrinsic motivation following the ad-ministration of tangibie rewards and increasesin intrinsic motivation following the adminis-tration of verbal rewards then is not surprising.

Behavioral Investigations of Intrinsic/Extrinsic Motivation

As Flora (1990) offered, "(behavioral)psychology is supposedly the study of indi-vidual behavior, not the study of groups means"

(p. 338). This statement succinctly illustratesthe importance of within-subject designs inbehavioral research. Behaviorally oriented re-searchers assert that cognitive researchers study-ing the effects of extrinsic reward using between-group designs have utilized measurement phasesthat are too short to detect temporal trends ortransition states (Cameron & Pierce, 1994). Inwithin-subject designs, however, behavior ismeasured over a number of sessions, therebyalleviating this shortcoming. Unlike between-group paradigms, within-subject design takesmeasurements over a number of sessions for eachphase. After baseline data are collected, rein-forcement is introduced and measurements areagain repeatedly taken. Finally, a withdrawal ofreinforcement occurs, and measurements of timeon task are taken again. Tune on task is oftenused as a measure of intrinsic motivation andthe difference in time on task between pre- andpostreinforcement is cataloged as intrinsic mo-tivation where differences are attributed to ex-ternal reinforcement. Behavioral investigationshave also traditionally included a follow-upphase during which measures of behavior aretaken 2 to 3 weeks after the conclusion of theexperiment to assess trends and temporal states.

Behaviorists' use of within-subject re-peated measures designs allows determinationof whether a reward is actually a reinforcer fora particular individual. Compared to the largenumber of group studies examining this sup-posed event, there are very few studies thatexamine the effects of extrinsic reinforcementfrom a behavioral standpoint (e.g., Akin-Little& Little, 2004; Davidson & Bucher, 1978;Feingold & Mahoney, 1975; Mawhinney et al.,1989; McGinnis, Friman, & Carlyon, 1999;Vasta, Andrews, McLaughlin, Stirpe, & Com-fort, 1978; Vasta & Stirpe, 1979). No studyutilizing this methodology has found evidenceof detrimental effect of the use of reward.

Akin-Little and Little (2004) attempted toexamnine the possible overjustification effects ofthe implementation of token economy for ap-propriate behavior. Although exhibiting appro-priate behavior in a classroom setting may notbe seen as intrinsically motivated behavior, manyreward contingency systems are used to in-crease compliant behavior. No previous studyused appropriate classroom behavior as the

352

Extrinsic Reinforcement in the Classroom

dependent variable although classroom manage-ment and student behavior is a major concern ofmany classroom teachers. The participants in thisstudy were 3 elementary school students chosenby their teacher as currently high in compliantbehavior to classroom rules. The token systemwas implemented by a regular classroom teacherin a classroom setting. Participants' behavior wasanalyzed after a Baseline I, reward procedure,Baseline II, and follow-up period. Nooverjustification effect was found for any of the3 students (i.e., no student's behavior droppedbelow Baseline I in either the Baseline II or fol-low-up phase) similar to the findings of Roane,Fisher, and McDonough (2003).

Using time on task playing a video gameas the dependent variable, Mawhinney andcolleagues (1989) attempted to differentiatebetween programmed and unprogrammed con-sequences and to ascertain whether or not ashift in locus of causality had occurred. Duringbaseline conditions participants were free tochoose to play video games or to manipulate atrigger pull task. In the reward phase where stu-dents were paid to play the video games withoutregard for total points scored, Mawhinney et al.posited that if playing the game came under thecontrol of extrinsic (programmed) rewards (i.e.,money), the students would play more games,but score fewer points because they would bemore concerned about finishing the game in theshortest amount of time to earn the most amountof money. On the other hand, if playing contin-ued to operate under the control of intrinsic(unprogramnmed) rewards, students' scores andtime playing the game would not be effectivelydifferent from baseline levels.

For 2 of the participants, Mawhinney etal. (1989) found that the amount of playing timeand points scored did not decrease with the in-troduction of reinforcement, indicating that theoffered reward did not result in a shift of loci ofcausality (i.e., the participants continued to playthe game "for its own sake"). Data from the thirdstudent indicated an increase in the number ofgames played, suggesting a shift in locus ofcausality from intrinsic to extrinsic rewards.However, although more games were played,the percentage of games won by this partici-pant remained approximately the same.

In their comparison of pie- andpostintervention data, Mawhinney et al. (1989)found that amount of quality of play (i.e., play-ing time and points scored) did not decreasefollowing the introduction of reinforcementacross participants. Participants continued toplay their preferred video game as much ormore during the postr,einforcement phase (i.e.,Baseline II) as during the prereinforcementphase (i.e., Baseline I). Therefore, the experi-menters concluded that providing money forthe performance of an interesting task did notdecrease intrinsic motivation regardless of theshift of locus of causality. The authors insteadsuggest an additive model as extrinsic rewardsincreased overall game performance.

Vasta and Stirpe (1979) asserted that theLepper et al. (1973) research generated twodiscrete courses of further research. One trendwas concerned primarily with the phenomenonthat undermined intrinsic motivation and uti-lized short-term, single-trial reward conditions.These researchers (Deci, 1971; Lepper et al.;Reiss & Sushinsky, 1975) explained the de-cline in play activity utilizing several differenttheories. Lepperetal. appliedBem's (1967) self-perception theory, whereby the introduction ofreinforcement adjusts the child's perception ofmotivation for the behavior from intrinsic toextrinsic, to elucidate their findings. Reiss andSushinsky suggested that decrements in behav-ior occur because of competing responses evokedby the introduction of the reward. Finally, Deci's(1975) cognitive evaluation theory has contin-ued to be applied to account for decreases in pre-scribed activities. Deci's theory asserts that areevaluation of motivation for one's behaviortakes place after extrinsic reinforcement that re-sults in a lowering of intrinsic motivation for theactivity. The second competing course of re-search focused on multiple-trial, long-term re-ward conditions (i.e., behavioral research meth-odology). In these studies, the reward is provento'be a reinforcer by its presentation over manysessions, not just one. The focus, then, is on theimplications Pfor intervention programs in theapplied setting, particularly thp school.

Because much of the difficulty in thisresearch rested in determining the definitionof intrinsic motivation, Vasta and Stirpe (1979)

353

School Psychology Review, 2004, Volume 33, No. 3

employed three dependent variables in theirstudy. Intrinsic motivation was defined as theamount of time a child engages in a certain task(Lepper et al., 1973), the amount of behaviorgenerated (Feingold &Mahoney, 1975), and thequality of performance (Kruglanski, Friedman,& Zeevi, 1971). Vasta and Stirpe measured theeffects of reinforcement on each of these depen-dent variables and found that receiving stars forcompletion of math worksheets functioned as areinforcer for the fourth grade school childrenin the experimental group. More importantly,however, the authors found no evidence of theundermining phenomenon on any of the threedependent variables used as measures of intrin-sic interest. Furthennore, the authors insisted thatthe recommendations by some researchers to de-crease the use of incentive programs because ofa possible undermining effect "appears unwar-ranted" (p. 241).

In an earlier study, Vasta et al. (1978)conducted two experiments. The participantswere 6 kindergarten/first grade students. Thefirst experiment purposefully chose partici-pants whose interest in the targeted activities(geo-blocks, cardboard puzzles, and copies ofpages from a coloring book) was low. The sec-ond experiment contained students whose ini-tial interest in the activities was relatively high.Both experiments contained four phases la-beled baseline, reinforcement, posttreatment,and follow-up. The coloring book activity wasrandomly chosen to be reinforced.

As with previous behavioral studies re-viewed, the data from these two experimentsindicated that the use of classroom-based rein-forcement programs was not detrimental tochildren's intrinsic interest in the activity. Themost important aspect of this study was Vasta etal.'s (1978) use of high and low interest groups.Lepper et al. (1973) pointed out that the use ofextrinsic reinforcementprograms was especiallydetrimental to children already displaying arela-tively high intrinsic interest in the activity. Thisassertion was not borne out by Vasta's findings,in that children in either the low or high initialinterest group did not display decreases in thenonreinforced response rate.

Feingold and Mahoney (1975) assessed5 second-grade children and attempted to test

the hypothesis that external reinforcement doesnot have a detrimental effect on children's in-trinsic interest in a play activity. Children weregiven Follow-the-Dots books and the optionof either connecting the dots to produce a pic-ture or playing with an Etch-A-Sketch. A re-peated measures, within-student design wasused and data were collected over a period offour phases: Baseline I, Reward, Baseline II,and Baseline m (i.e., Follow-Up phase, whichoccurred 2 weeks after Baseline II). Rewardsoffered included an assortment of candy, toys,and small books that children could win inexchange for points earned by connecting thedots in the Follow-the-Dots books. The rewardphase was confirmed to be a true reinforce-ment phase as it produced a dramatic increasein response for all children. Comparisons be-tween the last four sessions of Baseline I andBaseline II indicated no significant difference.Further, comparisons for each student weremade between Baseline I and pooled BaselineII and Im performances. Again, no significantdifferences were found, casting doubt on theassertion that external reinforcement has a del-eterious effect on the intrinsic interest of chil-dren in a play activity (Feingold & Mahoney).Finally, it should be noted that intrinsic moti-vation in both the cognitive studies (e.g., Deci,1971; Lepper et al. 1973) and most of the be-havioral studies (e.g., Davidson & Bucher,1978) was a measurement of time on task.

Recent Debate

Criticism of Cameron and Pierce (1994)

Although this debate originated withDeci's (1971) study, it gained impetus in 1994when Cameron and Pierce conducted a meta-analysis and concluded that reinforcement didnot harm an individual's intrinsic motivation.Subsequently, Cameron and Pierce's findingshave been criticized as utilizing flawed method-ology (Kohn, 1996; Lepper et al., 1996). Kohnwrote that Cameron and Pierce ignored impor-tant findings which suggested that the receipt oftangible rewards is associated with less volun-tary time on task as contrasted with the no-re-ward condition. Kohn further stated thatCameron and Pierce's methodology was flawed

354

Extrinsic Reinforcement in the Classroom

because results from studies in which infonna-tional praise was delivered (i.e., no detrimentaleffects on intrinsic motivation expected), withpraise delivered that might be construed as ma-nipulative (i.e., detrimental effects on intrinsicmotivation expected) were combined to detectan overall effect. Further, Kohn pointed out that,in his view, the more common type of praise ina classroom is the latter, and, therefore, studiesthat utilized manipulative praise should be ex-amined separately. Kohn continued his criticismagainst token reinforcement programs by in-specting studies examining the effects of per-fonnance-contingent rewards (PCRs). Kohn re-futed the idea that the delivery of PCRs can miti-gate the detrimental effect on children's intrin-sic motivation by stating that in the majority ofprograms in "real" classrooms, not all childrenattain the specified criterion level and still ob-tain rewards. This is in opposition to researchstudies in which variables are manipulated sothat each participant receives a reward. There-fore, Kohn suggested that a very different moti-vational effect might be expected in an actualclassroom. Kohn concluded by expressing hispersonal belief that adequate justification existsfor schools to avoid the use of incentive pro-grams and simply provide children with infor-mational feedback. This informational feedbackalone is expected to increase compliant behav-ior and work output, and maintain initial lev-els of intrinsic motivation.

Lepper et al. (1996) characterizedCameron and Pierce's (1994) meta-analysis asbeing overly simplistic and of little theoreticalvalue. They claimed that Cameron and Piercemisused meta-analytic techniques, and attemptedto offer amore nuanced analysis. Similar to Kohn(1996), Lepper et al. stated that the 1994 meta-analysis should have focused on the possiblefunctions of rewards and the possible detrimen-tal implications (i.e., a reward can serve an in-strumental/incentive function, an evaluation/feedback function, or function as a social con-straint) instead of an overall effect.

. Cameron and Pierce (1996) respondedto these criticisms by first stating that investi-gating the overall effect of extrinsic rewards isnecessary for practical and theoretical reasons.From a practical standpoint, it is clear that

many parents, educators, and administratorshave embraced Kohn's (1993) view that over-all, incentive systems are damaging. Manyclassroom teachers, however, still wish to adoptan incentive program. These teachers are,therefore, interested in whether or not, over-all, rewards disrupt intrinsic motivation forcompleting work or attaining a specified levelof performance. The overall effect of reward,then, is critical to educational strategy(Cameron & Pierce). Many academic joumalsand textbooks point to the theoretical overalldetrimental effects of rewards or reinforcement.Consequently, many parents, teachers, and oth-ers are loath to use any reinforcement procedureunder any conditions. It is necessary then, ac-cording to Cameron and Pierce, to analyze theoverall effect of rewards because many writersare criticizing the use of incentive programs ineducational settings. These criticisms are basedupon research findings that some interpret asindicating an overall negative effect. Cameronand Pierce concluded their response by statingthat their meta-analysis is the most thorough todate in comparison to other meta-analyses onthis topic, specifically Tang and Hall's (1995)analysis that included 50 studies; Wiersma's(1992) analysis that contained 20 studies; andRummel and Feinberg's (1988) analysis thatcomprised 45 studies. Each of these analysesreported overall that extrinsic rewards had det-rimental effects on intrinsic motivation. Thesefindings were in direct contrast to the conclu-sions of Cameron and Pierce who stated emphati-cally that their results, from an analysis of over100 studies that included all of the relevant stud-ies in this research area, illustrated that rewardscan be utilized to maintain or even enhance in-trinsic motivation. Further, and more importantly,the conditions under which detrimental effectsto intrinsic motivation are exhibited occur un-der highly circumscribed conditions (e.g., re-wards delivered with no set criterion), situa-tions that are easily eschewed by the properuse of token reinforcement programs.

Deci et al. (1999a) and Cameron et al.(2001)

In response to Cameron and Pierce's(1994) meta-analytic findings, Deci et al.

355

School Psychology Review, 2004, Volume 33, No. 3

(1999a) conducted a separate meta-analysis toprovide a direct test of cognitive evaluationtheory. A total of 128 studies were analyzed andseparate effect sizes were calculated. The re-searchers specifically examined verbal rewards(termed "positive feedback") separately fromtangible rewards. In addition, further analysis ofstudies incorporating tangible rewards was con-ducted, wherein separate effect sizes were com-puted for unexpected and expected, tasknoncontingent (i.e., rewards given not for en-gaging in the task specifically, but for participa-tion in the experiment), engagement contingent(i.e., rewards given for participation in the task),completion contingent (i.e., rewards given forcompletion of the task), and performance con-tingent (i.e., rewards given only for performingthe task well, or surpassing a previously set stan-dard). Furthermore, the results were dividedbased on participant population (i.e., child ver-sus college student). Not surprisingly, a decre-ment in intrinsic motivation, measured by timeon task for 101 of the studies and self-report ofinterest for 84 of the studies, was found in everycategory except verbal rewards and unexpectedrewards. Deci and colleagues (2001) argued thatthe results of their meta-analysis provide sup-port for cognitive evaluation theory and confirmthe substantial undermining effects following theuse of external rewards. Deci et al. evinced theimportance of considering the interpersonal con-text in the delivery of verbal reward specificaly(i.e., rewards delivered in a controling mannerwill tend to decrease intrinsic motivation whereasrewards delivered in a noncontrolling mannerwil tend to increase feelings of competence and,hence, intrinsic motivation).

Although verbal rewards enhanced theintrinsic motivation of college students (i.e., sig-nificant increase), the delivery of verbal rewardsdid not enhance children's intrinsic motivation.Based on these findings, Deci et al. (2001) cau-tioned about the use of verbal reward in the class-room: "verbal rewards are less likely to have apositive effect for children ... (they) can evenhave a negative effect on intrinsic motivation"(p. 9). This is misleading. The importance andeffectiveness of teacher attention particularly inthe form of verbal praise has been documented(Drevno et al., 1994; Maag & Katsiyannis,

1999; Parrish, Cataldo, Kolko, Neef, & Egel,1986; Valcante, Roberson, Reid, & WoDitng,1989). The assertion that verbal praise shouldnot be utilized in a classroom setting is in di-rect opposition to the available data.

Deci et al. (1999a, 2001) examined ageeffects on tangible rewards in the task engage-ment category. They found that although bothcollege students and children experience dec-rements in intrinsic motivation after the use ofexternal reward, it is worse for children. Basedon these findings, they wrote:

Tangible rewards, both material, such aspizza parties for reading books, and symbolicrewards, such as good student awards arewidely advocated by many educators and areused in many classrooms, yet the evidencesuggests that these rewards tend to under-mine intrinsic motivation for the rewardedactivity. Because the undermining of intrin-sic motivation by tangible rewards was es-pecially strong for school-agedchildren... .the findings from this meta-analysis are of particular import for primaryand secondary school educators. (Deci et al.,2001, p. 15)

This finding is in direct contrast to a study con-ducted by Flora and Flora (1999) who exam-ined the reading habits of college students whoreported having been rewarded for reading asan elementary school student by participating inthe "Booklt" program. No detrimental effect wasfound on college students' reading habits after hav-ing been rewarded for reading as a youngster.

In the most recent meta-analysis,Cameron et al. (2001) synthesized 145 studiesusing categorizations similar to those adoptedby Deci et al. (1999a). They found, in general,that rewards do not decrease intrinsic motiva-tion. Although the sample was not homoge-neous, an overall effect size was calculated.Cameron (2001) stated this overall effect isimportant because educators and other schoolpersonnel often report that all rewards areharmful on motivation. Contrary to Deci et al.(1999a), Cameron et al. included the catego-ries of high and low initial interest. Notably,they found that rewards can enhance intrinsicmotivation, particularly if measured as time ontask. This is in accordance with Bandura's(1986) finding that most activities have littleinitial interest for people, but that engagementin the activity may increase interest. This has

356

Extrinsic Reinforcement in the Classroom

important implications for schools as many chil-dren do not find academic tasks initially appeal-ing. The use of reward then may be used to in-crease students' time on task and intrinsic moti-vation for a task. Cameron et al. did not finddetrimental effects with the use of verbal praisefor either children or college students. Instead,they found significant increases.

In terms of tangible reward, no detrimen-tal effect was found for unexpected rewards orfor rewards that were closely tied to specific stan-dards of performance and to success. Detrimen-tal effect was found when rewards were not ex-plicitly connected to the task and signified fail-ure. This last finding is also important to educa-tors who may be attempting to use reinforce-ment to increase either social or academic be-havior. Oftentimes, teachers will set the goalsfor a student too high. Behavioral principles statethat it is important to shape behavior, reinforc-ing the child's current competencies and giv-ing him or her a chance for success.

It is also important to remember that nei-ther of these meta-analyses examined the resultsof behavioral studies (e.g., Feingold & Mahoney,1975). Again, none of the studies that utilizedsingle case design found any detrimental effectswith the use of reinforcement contingencies. Thisis significant as those studies tend to more typi-cally mimic the use of reward contingencies inclassrooms. Perhaps if additional behaviorallyoriented studies were conducted, there would beno detection of the supposed detrimental effectsof reward delivery on any task or behavior.

Best Practices in the Use ofReinforcementProcedures in the

Classroom

In 1991, the National Education Asso-ciation published a document entitled How toKill Creativity (Tegano et al.) that stated:

The expectation of reward can actually un-dermine intrinsic motivation and creativityof performance... .A wide variety of rewardshave now been tested, and everything fromgood-player awards to marshmallows pro-duces the expected decrements in intrinsicmotivation and creativity of performance....(making) them (students) much less likelyto take risks or to approach a task with a play-ful or experimental attitude. (p. 119)

A review of several educational psychologybooks (Slavin, 1997; Woolfolk, 2004) reveals amore balanced view of the effects of rewards byincluding the findings of Cameron and Pierce(1994), along with Deci and Ryan (1985) andLepper et al. (1973). This is an encouraging signbecause many of the findings in this area sup-port the effectiveness of reinforcement proce-dures in the classroom and many researchershave criticized the literature on supposed dam-aging effects (e.g., Bandura, 1986; Bates, 1979;Dickinson, 1989; Flora, 1990).

Additionally, any detrimental effects of theuse of extrinsic reinforcement can be easilyavoided with the use of these guidelines. Re-wards should not be presented for mere partici-pation in a task without regard for completionor quality. Decrements have also been found inthe literature when rewards are presented on asingle occasion. This is not the most commonmethod utilized in classrooms. In general, re-ward contingencies used in schools are presentedrepeatedly with appropriate thinning of sched-ules utilized when behavior change has occurred.School psychologists are advised to heed this ad-vice when consulting and planning with teacherson the use of reinforcers in the school setting.

Specifically, school psychologists areoften asked to aid in increasing the frequency ofa number of student behaviors (e.g., math, read-ing, homework) for which the baseline level ofperformance is close to zero (i.e., the "unmoti-vated" child). Maintaining a perspective thatthese students "should" engage in certain behav-iors because of "intrinsic" motivation is unlikelyto result in a change in the level of performance.Instead, the efficacious response includes select-ing the target behavior(s), determining the cur-rent and desired level of functioning, and deliv-ering reinforcers based on a set criterion. Thisciiterion changes as the behavior improves. Thisentire procedure is based on the principles ofshaping through reinforcement of successive ap-proximations of the desired behavior. This prac-tice has been used with both performance andacquisition deficits, with acquisition deficits re-quiring a slightly different schedule (i.e., con-tinuous) of reinforcer delivery in the initiallearning stages. Additionally, to insure thatextrinsic rewards have true reinforcing value,

357

School Psychology Review, 2004, Volume 33, No. 3

the use of a Reinforcer Preference Survey(Northup, George, Jones, Broussard, &Vollmer, 1996) has proven efficacious.

Teachers continually request training inbehavior and classroom management techniques(Maag, 1999,2001) such as the procedures dis-cussed above. The irony is that techniques thataid teachers in improving their managementskills have existed since Skinner's (1953) semi-nal work on the principles of operant condition-ing. Techniques based upon the use of extrinsicreinforcers (i.e., positive reinforcement) work inthe classroom. These include verbal praise, to-ken economies, group contingencies, contracts,and others (Maag & Kotlash, 1994). The ques-tion is why teacher education programs are notincorporating these principles into their curricu-lum. Why is there such resistance to the data?Axelrod (1996) suggested that some causes forthe lack of both professional and popular accept-ability (e.g., Kohn, 1993) may be that the use ofpositive reinforcement consumes too much time,attempts to eliminate human choice, and thereis little compensation for educational personnelfor using these procedures. This is a somewhatdiscouraging view and one can only hope thatfuture and current teachers, educational person-nel, and teacher training faculty make evidence-based decisions when choosing intervention forchildren and youth.

The polemical papers on both sides ofthis extrinsic/intrinsic issue, their rebuttals, andthe further replies are gradually bearing out acommon "bottom line"-the programs thatshow increased intrinsic motivation are thoseprograms that incorporate the elements ofgood, comprehensive behavioral intervention:relatively immediate reinforcement, generali-zation strategies, and individualized interven-tion. The implication is that any blanket rejec-tion of programmed reinforcement strategiesis entirely unwarranted and programmed re-inforcement strategies, like any other instruc-tional strategy, should be undertaken in athoughtful manner after considering the manyvariables of any classroom situation.

Future Research

There are a number of areas where futureresearch is warranted. First, research should be

designed that attempts to inspect the Lepper etal. (1996) hypothesis that intrinsic motivation isstrengthened under certain extrinsic reward con-ditions (e.g., reward indicates success and in-creased feeling of competence) or intrinsic mo-tivation is decreased under less favorable extrin-sic reward conditions (e.g., constraints or socialcontrol function). Studies in this area wiUl needto focus on the conditions under which behav-ioral decrements are a result of decreases in "in-trinsic" motivation or environmental factors(e.g., schedules of reinforcement delivery, con-trolling for the value of different rewards, andbehavioral contrast). As Carton (1996) wrote,"choosing between an explanation based on in-trinsic motivation versus one based on environ-mental variables is a decision that can lead tovery different conclusions" (p. 247).

Another promising area is the re-exami-nation of seminal studies (e.g., Deci, 1971;Lepper et al., 1973) using both the originalmethodology and a repeated measures design.The first and fourth author are currently in-volved in a project that is attempting to rep-licate the Lepper et al. study. At the conclu-sion, that same study (i.e., observing chil-dren who currently draw, rewarding them,and then observing their behavior post-reward)will be conducted utilizing a repeated mea-sures design including a follow-up phase.Future research should attempt to ascertainthe validity of theories such as the cognitiveevaluation theory and the overjustification ef-fect using methodology similar to that de-scribed above. Only through an examinationof behavior across time can the true effects ofthe use of reinforcers be determined.

Intrinsic motivation has been defined asbehaviors performed in the absence of observ-able extemal reinforcement. Defining any con-struct in terms of what it is not does little toadvance the course of science. Further studiesmay eventualy show the necessity of examin-ing whether or not intrinsic motivation is auseful concept to study or whether it even ex-ists. There is an extensive literature support-ing the efficacy of behavioral interventions ina variety of settings. Perhaps the time has cometo accept these findings and cease attempts todamage these data by offering altemative, in-

Extrinsic Reinforcement in the Classroom

validated conclusions. Perhaps it will then bepossible to avoid Skinner's (1953) epistemologi-cal criticism of hypothesized inner causes. Nolonger will we be content with attributing be-havior to "intrinsic motivation," but we willsearch for the environmental factors that are theultimate causes of behavior. In this light, thedebate over detrimental effects of reward can beseen as being due in large part to confusion overvalid candidates for causes.

Conclusion

Behavior analysts have spent their timeexploring the variables that affect the efficacyof reinforcement. They have always admittedthat under certain circumstances reinforcementis more effective (e.g., in the presence of an es-tablishing operation; Michael, 1993) or less ef-fective (e.g., when discriminative stimuli are notpresent to signal the availability of reinforce-ment). Cognitive researchers have focused onthose instances when reinforcement is less ef-fective, and attributed this lack of efficacy to ageneral problem with programmed reinforce-ment, claiming that such reinforcement alters aninner propensity called intrinsic motivation. Inexamining the methodology of these experi-ments, however, it often becomes clear that thereinforcement programs are simply bad pro-granms-that is, they do not exploit those strate-gies that we know make for effective reinforce-ment programs. When tangible rewards are notdelivered immediately after behavior, when anindividual's baseline performance is not takeninto account in intervention design (as in whenstudents who are already performing a task at ahigh frequency are put on programmed reinforce-ment), and when generalization strategies are notused, it is hardly surprising that "intrinsic moti-vation" is lowered. The logical solution is not toeliminate programmed reinforcement, but to useeffective programmed reinforcement strategies.

Bribery is defined in the dictionary as aninducement to engage in illegal or inappropriatebehavior (Woolf, 1980). When education per-sonnel, including school psychologists, extol theuse of extrinsic reinforcement in the classroom,the motive is clearly not to "bribe'? children andyouth, but to increase appropriate academic andsocial behavibr. The goal is obviously not to

decrease intrinsic motivation, although it is un-clear that the construct exists or is useful in thescience of psychology. It is apparent through anexamination of the data that any decrease oc-curs under specifically circumscribed conditions,conditions that are easily avoidable. Best prac-tice would suggest that children and youth de-serve interventions based on sound, empiricalfindings. The positive effect of the use of rein-forcers in the classroom is one such conclusion.

If teachers are implementing "reinforce-ment" programs without knowing how to do so,the worries of intrinsic motivation researchersseem reasonable. But it is the practice and notthe principle that is suspect and open to misap-plication and abuse, and the corresponding pro-phylactic is more teacher training in behavioralmethods, not less. In the meantime, we can onlytry to correct the misconceptions that have ledto unwarranted criticism of programmed rein-forcement, and take some solace in the fact thatclassroom teachers will continue to leam fromcontingent consequences what works and whatdoes not.

References

Akin-Little, K. A., & Little, S. G. (2004). Re-examiningthe oveijustification effect. Journal ofBehavioral Edu-cation, 13, 179-192.

Alberto, P. A., & Troutman, A. C. (2003). Applied behav-ior analysis for teachers (6'aed.). Upper Saddle River,NJ: Prentice Hall.

Axelrod, S. (1996). What's wrong with behavior analy-sis? Journal of Behavioral Education, 6, 247-256.

Ayllon, T., & Azrin, N. (1968). The token economy: Amotivational system for therapy and rehabilitation.New York: Appleton-Century-Crofts.

Bandura, A. (1977). Social learning theory. EnglewoodCliffs, NJ: Prentice Hall.

Bandura, A. (1986). Social foundations of thought andaction: A social cognitive theory. Englewood Cliffs,NJ: Prentice Hall.

Banish, H. H., Saunders, M., & Wolf, M. M. (1969). Goodbehavior game: Effects of individual contingencies forgroup consequences on disruptive behavior in a class-room. Journal ofApplied BehaviorAnalysis, 2, 119-124.

Bates, J. A. (1979). Extrinsic reward and intrinsic motiva-tion: A review with implications for the classroom.Review of Educational Research, 49, 557-576.

Bem, D. J. (1967). Self-perception: An alternative inter-pretation of cognitive dissonance phenomena. Psycho-logical Review, 74, 183-200.

Bernstein, D. J. (1990). Of carrots and sticks:Areview ofDeci and Ryan's Intrinsic Motivation and Self-Deter-mination in Human Behavior. Journal of Experimen-tal Analysis of Behavior, 54, 323-332.

359

School Psychology Review, 2004, Volume 33, No. 3

Bruner, J. S. (1960). The process of education. Cambridge,MA: Harvard University Press.

Buisson, G. J., Murdock, J.Y., Reynolds, K. E., & Cronin,M. E. (1995). Effect of tokens on response latency ofstudents with hearing impairments in a resource room.Education and Treatment of Children, 18, 408-421.

Cameron, J. (2001). Negative effects of reward on intrin-sic motivation-A limited phenomenon: Comment onDeci, Koestner, and Ryan (2001). Review of Educa-tional Research, 71, 29-42.

Cameron,J., Banko, K. M., & Pierce, W. D. (2001). Perva-sive negative effects of rewards on intrinsic motivation:The myth continues. The Behavior Analyst, 24, 1-44.

Cameron, J., & Pierce, W. D. (1994). Reinforcement, re-ward, and intrinsic motivation: A meta-analysis. Re-view of Educational Research, 64, 363-423.

Cameron, J., & Pierce, W. D. (1996). The debate aboutrewards and intrinsic motivation: Protests and accusa-tions do not alter the results. Review of EducationalResearch, 66, 39-51.

Carton, J. S. (1996). The differential effects of tangiblerewards and praise on intrinsic motivation: A compari-son of cognitive evaluation theory and operant theory.The BehaviorAnalyst, 19, 237-255.

Cavalier, A. R, Ferretti, R. P., & Hodges, A. E. (1997).Self-management within a classroom token economyfor students with leaming disabilities. Research inDevelopmental Disabilities, 18, 167-178.

Davidson, P., & Bucher, B. (1978). Intrinsic interest andextrinsic reward: The effects of a continuing token pro-gram on continuing nonconstrained preference. Behav-ior Therapy, 9, 222-234.

Davis, R. T., Settlage, P. H., & Harlow, H. F. (1950). Per-formance of normal and brain-operated monkeys onmechanical puzzles with and without food incentives.Journal of Genetic Psychology, 7, 305-311.

DeCharms, R. (1968). Personal causation: The internalaffective determinants of behavior New York: Aca-demic Press.

Deci, E. L. (1971). Effects of externallymediatedrewardson intrinsic motivation. Journal of Personality andSocial Psychology, 18, 105-115.

Deci, E. L. (1975). Intrinsic motivation. New York: Ple-num Press.

Deci, E. L., Koestner, R., & Ryan, R. M. (1999a). Ameta-analytic review of experiments examining the effectsof extrinsic rewards on intrinsic motivation. Psycho-logical Bulletin, 125, 627-668.

Deci, E. L., Koestner, R., & Ryan, R. M. (1999b). The un-dermining effect is a reality after all-extrinsic rewards,task interest, and self-determination. Reply toEisenberger, Pierce, and Cameron (1999) and Lepper,Henderlong, and Gingras (1999). Psychological Bulle-tin, 125, 692-700.

Deci, E. L., Koestner, R., & Ryan, R. M. (2001). Extrinsicrewards and intrinsic motivation in education: Recon-sidered once again. Review of Educational Research,71, 1-27.

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

Dickinson, A. M. (1989). The detrimental effects of ex-trinsic reinforcement on "intrinsic motivation." TheBehaviorAnalyst, 12, 1-15.

Drevno, G., Kimball, J., Possi, M., Heward, W., Gardner,R., & Barbetta, P. (1994). Effects of active student re-sponse during error correction on the acquisition, main-tenance, and generalization of science vocabulary byelementary students: A systematic replication. Jour-nal ofApplied BehaviorAnalysis, 27, 179-180.

Eisenberger, R., & Cameron, J. (1996). Detrimental ef-fects of reward: Reality or myth? American Psycholo-gist, 51, 1153-1166.

Eisenberger, R., Pierce, W. D., & Cameron, J. (1999). Ef-fects of reward on intrinsic motivation: Negative, neu-tral, and positive. Psychological Bulletin, 125, 677-691.

Fair, E. M., & Silvestri, L. (1992). Effects of reward, com-petition and outcome on intrinsic motivation, Journalof Instructional Psychology, 19, 3-8.

Feingold, B. D., &Mahoney, M. J. (1975). Reinforcementeffects on intrinsic interest: Undermining theoverjustification hypothesis.BehaviorTherapy, 6,367-377.

Flora, S. R. (1990). Undermining intrinsic interest fromthe standpoint of a behaviorist. The PsychologicalRecord, 40, 323-346.

Flora, S. R., & Flora, D. B. (1999). Effects of extrinsicreinforcement for reading during childhood on reportedreading habits of college students. The PsychologicalRecord, 49, 3-14.

Greene, D., & Lepper, M. R. (1974). Effects of extrinsicrewards on children's subsequent intrinsic interest.Child Development, 45, 1141-1145.

Guzzo, R. A. (1979). Types of rewards, cognitions andwork motivation. Academy of Management Journal,22, 75-86.

Harlow, H. F., Harlow, M. K., & Meyer, D. R. (1950).Learning motivated by manipulative drive. Journal ofExperimental Psychology, 40, 228-234.

Horcones, C. (1987). The conceptof consequences in the analy-sis of behavior. The BehaviorAnalyst, 10, 291-294.

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

Kohn, A. (1996). By all available means: Cameron andPierce's defense of extrinsic motivators. Review ofEducational Research, 66, 1-3.

Kruglansld, A. W., Friedman, I., & Zeevi, G. (1971). The ef-fects of extrinsic incentive on some qualitative aspects oftask performance. Journal of Personality, 39, 606-617.

Lepper, M. R. (1983). Extrinsic reward and intrinsic mo-tivation: Implications forthe classroom. In J. M. Levine& M. C. Wang (Eds.), Teacher and student percep-tions: Implications for learning (pp. 281-317).Hillsdale, NJ: Erlbaum.

Lepper, M. R. (1998). Awhole much less-than the sum ofits parts. American Psychologist, 53, 675-676.

Lepper, M. R., & Gilovich, T. (1981). The multiple func-tions of reward: A social developmental perspective.In S. S. Brehm, S. Kassin, & F. X. Gibbons (Eds.),Developmental social psychology (pp. 5-31). NewYork: Oxford University Press.

Lepper, M. R., Greene D., & Nisbett, R. E. (1973). Under-mnining children's intrinsic interest with extrinsic reward:A test of the "overjustification" hypothesis. Journal ofPersonality and Social Psychology, 28, 129-13 7.

Lepper, M. R., Keavney, M., & Drake, M. (1996). Intrin-sic motivation and extrinsic rewards: A commentary

360

Extrinsic Reinforcement in the Classroom

on Cameron and Pierce's meta-analysis. Review ofEducational Research, 66, 5-32.

Maag, J. W. (1996). Parenting without punishment. Phila-delphia: The Charles Press.

Maag, J. W. (1999). Behavior management: From the theo-retical implications to practical applications. San Di-ego: Singular.

Maag, J. W. (2001). Rewarded by punishment: Reflec-tions on the disuse of positive reinforcement in schools.Exceptional Children, 67, 173-186.

Maag, J. W., & Katsiyannis, A. (1999). Teacher prepara-tion in E/BD: A national survey. Behavioral Disorders,24, 189-196.

Maag, J. W., & Kotlash, J. (1994). Review of stress in-oculation training with children and adolescents: Is-sues and recommendations. BehaviorModification, 18,443-469.

Mawhinney, T. C., Dickinson, A. M., & Taylor, L. A., HI.(1989). The use of concurrent schedules to evaluatethe effects of extrinsic rewards on "intrinsic motiva-tion." Journal of Organizational Behavior Manage-ment, 10, 109-129.

McClelland, D. C.,Atkinson, J. W., Clark, R. W., & Lowell,E. L. (1953). The achievement motive. New York:Appleton-Century-Crofts.

McGinnis, J. C. (1996). On intrinsic motivation: A pro-posalfor and validation of a unique dependent mea-sure. Unpublished masters thesis, University of South-ern Mississippi, Hattiesburg.

McGinnis, J. C., Friman, P. C., & Carlyon, W. D. (1999).The effect of token rewards on "intrinsic" motivationfor doing math. Journal ofApplied BehaviorAnalysis,32, 375-379.

Michael, J. (1993). Establishing operations. The Behav-iorAnalyst, 16, 191-206.

Morgan, M. (1984). Reward-induced decrements and in-crements in intrinsic motivation. Review of EducationalResearch, 54,5-30.

Northup, J., George, T., Jones, K., Broussard, C., &Vollmer, T. R. (1996). A comparison of reinforcer as-sessment methods: The utility of verbal and pictorialchoice procedures. Joumal ofApplied BehaviorAnaly-sis, 29, 201-212.

O'Leary, K. D., & Drabman, R. (1971). Token reinforce-ment programs in the classroom: A review. Psycho-logical Bulletin, 75, 379-398.

Parrish, J., Cataldo, M., Kolko, D., Neef, N., & Egel, A.(1986). Experimental analysis of response covariationamong compliant and inappropriate behavior. Journalof Applied BehaviorAnalysis, 19, 241-254.

Pittman, T. S., Boggiano, A. K., & Ruble, D. N. (1983).Intrinsic and extrinsic motivational orientations: Lim-iting conditions on the undermining and enhancingeffects of reward on intrinsic motivation. In J. M.Levine & M. C. Wang (Eds.), Teacher and studentperceptions: Implications for learning (pp. 319-340).Hillsdale, NJ: Erlbaum.

Reiss, S., & Sushinsky, L. W. (1975). Overjustification,competing, responses, and the acquisition of intrinsicinterest. Journal of Personality and Social Psychol-ogy, 30, 1116-1125.

Roane, H. S., Fisher, W. W., & McDonough, E. M. (2003).Progressing from programmatic to discovery research:A case example with the overjustification effect. Jour-nal ofApplied BehaviorAnalysis, 36, 35-46.

Rosenfield, D., Folger, R., &Adelman, H. F (1980). Whenrewards reflect competence: A qualification of theoverjustification effect. Journal of Personality andSocial Psychology, 39, 368-376.

Rummel, A., & Feinberg, R. (1988). Cognitive evaluationtheory: A meta-analysis review of the literature. So-cial Behavior and Personality, 16, 147-164.

Ryan, R. M., & Deci, E. L. (1996). When paradigms clash:Comments on Cameron and Pierce's claim that rewardsdo not undermine intrinsic motivation. Review of Edu-cational Research, 66, 33-38.

Scott, W. E., Jr. (1975). The effects of extrinsic rewardson "intrinsic motivation." Organizational Behavior andHuman Performance, 25, 311-335.

Skinner, B. F (1953). Science and human behavior. NewYork: Macmillan.

Slavin, R. E. (1997). Educational psychology (5" ed.).Needham Heights, MA: Allyn & Bacon.

Swiezy, N. B.: Matson, S. L., & Box, P. (1992). The goodbehavior game: A token reinforcement system forpreschoolers. Child and FamilyBehavior Therapy, 14,21-32.

Tang, S., & Hall, V. (1995). The oveijustification effect: Ameta-analysis. Applied Cognitive Psychology, 9, 365-404.

Tegano, D. W., Moran, D. J., III, & Sawyers, J. K. (1991).Creativity in early childhood classrooms. Washington,DC: National Education Association.

Valcante, G., Roberson, W., Reid, W., & Wolking, W.(1989). Effects of wait-time and intertrial interval du-rations on leaming by children with multiple handi-caps. Journal of Applied Behavior Analysis, 13, 43-55.

Vasta, R., Andrews, D. E., McLaughlin, A. M., Stirpe, L.A., & Comfort, C. (1978). Reinforcement effects onintrinsic interest: A classroom analog. Journal of SchoolPsychology, 16, 161-166.

Vasta, R., & Stirpe, L. A. (1979). Reinforcement effectson three measures of children's interest in math. Be-havior Modification, 3, 223-244.

White, R. W. (1959). Motivation reconsidered: The con-cept of competence. Psychological Review, 66, 297-333.

Wiersma, U. J. (1992). The effects of extrinsic rewards inintrinsic motivation: Ameta-analysis. Journal of Occu-pational and Organizational Psychology, 65, 101-114.

Williams, B. W. (1980). Reinforcement, behavior con-straint and the overjustification effect. Journal of Per-sonality and Social Psychology, 39, 599-614.

Woolf, H. B. (1980). Webster's new collegiate dictionary.Springfield, MA: G. & C. Merriam Company.

Woolfolk, A. (2004). Educational psychology (9" ed.).Boston: Allyn & Bacon.

Zimmerman, B. J. (1985). The development of "intrinsic"motivation: A social leaming analysis. Annals of ChildDevelopment, 2, 117-160.

361

School Psychology Review, 2004, Volume 33, No. 3

362

K. Angeleque Akin-Little, PhD, is an Assistant Professor of School Psychology at theUniversity of the Pacific. Her primary interests include applied behavior analysis, par-ticularly the effects of extrinsic reinforcement on intrinsic motivation, professional is-sues, overscheduling, and systems change. She graduated from The University of South-ern Mississippi.

Tanya L. Eckert, PhD, is an Associate Professor of Psychology at Syracuse University.Her research interests include examining procedures for assessing academic and behav-ioral problems, developing classroom-based interventions, and measuring the accept-ability of assessment and intervention procedures. She earned her PhD from LehighUniversity.

Benjamin J. Lovett is a doctoral candidate in School Psychology at Syracuse University.

Steven G. Little, PhD, is Professor and Chair of the Department of Educational andSchool Psychology at the University of the Pacific in Stockton, California. He has previ-ously held academic positions in Alabama, Elinois, and New York. His research interestsinclude teacher attributions for student behavior, applied behavior analysis, overschedulingof children and youth, and professional issues in school psychology. He received his PhDin School Psychology from Tulane University.

COPYRIGHT INFORMATION

TITLE: Extrinsic Reinforcement in the Classroom: Bribery orBest Practice

SOURCE: Sch Psychol Rev 33 no3 2004WN: 0400300759003

The magazine publisher is the copyright holder of this article and itis reproduced with permission. Further reproduction of this article inviolation of the copyright is prohibited. To contact the publisher:http://www.nasponline.org/index2.html

Copyright 1982-2004 The H.W. Wilson Company. All rights reserved.