university of illinois library at urbana-champaigm Stacks Digitized by the Internet Archive in 2011 with funding from University of Illinois Urbana-Champaign http://www.archive.org/details/selfperceptionof89cald Faculty Working Papers ^H THE ^^1 EXTRIN ^ ^H SELF-PERCEmON OF INTRINSIC AND EXTRINSIC MOTIVATION; A CRITICAL REVIEW Bobby J. Calder and Barry M. Staw ,^^89 College of Commerce and Business Administration University of Illinois at U rb a n a - C h a m p a i g n FACULTY WORKING PAPERS College of Commerce and Business Administration University of Illinois at Urbana -Champaign February 22, 1973 THE SELF-PERCEimON OF INTRINSIC AND EXTRINSIC MOTIVATION; A CRITICAL REVIEW Bobby J. Calder and Barry M. Staw The Self-Perc'jption of Intrinsic and Extrinsic Kotivationi A Critical Review Botby J, Calder and 3arry M, 3tav? University of Illinois at Ur bana-Chcun {»ai gn Self -Perception and motivation Abstract A distinction is comnonly made between intrinsic and extrinsic motiva- tion. It is argued here that, although the status of intrinsic motivation as an actual psychological process is unclear, it may be useful to investigate the self-percepticn of intrinsic versus extrinsic motivation, de Charms, for excuaple, has hypothesized that increasing extrinsic rewards raay lead individuals to perceive their beha^/ior as under the control of these rewards eind not intrinsically motivated. The consequences of such perceptions for organizational behavior are discussed. The purpose of this review is to sharpen some of the theoreticaJ. issues at stake in a self-perception approach to motivation and to critically review several relevant studies in order to direct future research. The Self -Perception of Intrinsic Motivation! A Critical Review Resea.rch on motivation has frequently drawn a distinction between intrinsic atnd extrinsic iTiOtivation (e.g., Atkinson, 1964; Hunt, 1965{ Koch, 1956{ Youag« 1961). If a situation contains a specific ,^oal which provides satisfawjxxon independent of the actual acti'lty itself, behavior is said to be extrlnsically motivated. On the other hand, if the activity is valued for its own sake auid appears to be self -sustained, behavior is said to be intrinsically motivated (Young, I96I, p. 171)^ An early description by Woodworth (1918) captures the phenomenon vielli . . .While a man may enter a certain line of business ftom a purely external economic motive, he develops am interest in the business for its own sake, . , and the rotiv;. force that drives him in the daily task, provided of course this does not degenerate into mere automatic routine, is precisely an interest in the problems confronting him and in the processes by wliich he is able to dsal with those problems, 2^-® 6A^ fumish &s th_e motiv e force for the search for means but o nce the means are found , they are agt_ to becof.iq interestin.cc c£ their own account Ti tallies added, p. iOil'J. Althoujh this dislinotio":. is conceptUi'.lly appealing, it raises difficult questions. The purpose of the present discussion of intrinsic motivation is to iinarpen some of the theoretical issues at stake and to evaluate the raethodolcv^ca.1 adequa,cy of a number of recent experiments. There are two major problem.-; confronting the account of any behavior in terms of intrinsic and extrinsic motivation (Cofer and Appley, I967), The most serious is that the phenomenon is merely named, not expleiined. Labeling a behavior as intrinsically motivated begs the question of the theoretical nature of the process through vMch the behavior has become a motive. The seconc'. problem is that there are other theories which -2- might plausibly explain the phenomenon. No doubt the most common alter- native explanation involves secondary reinforcement. Secondary reinforce- ment refers to a process by vlnlch an orj.ginally neutral stimulus acquires reinforcing properties through its association with a primary reinforcer. In these terms, an intrinsically motivated activity is simply one in which the reinforcement value of the goal has associatively rubbed off on the behavior itself. It is difficult then to use the notion of intrinsic motivation beyond the descriptive level. While the notion may fit our intuitions about labeling various motives, this does not constitute a psychological explanation of behavior. Researchers in the area of organizational behavior seem to have implicitly recognized this problem by employing intrinsic motivation mainly to denote a certain class of motives. Herz berg's (1966) motivational approach, for instauice, literally postulates both intrinsic factors (e.g., recognition and achievement) and extrinsic factors (e.g., pay and working conditions) as determinants of job behavior. Any particular factor is simply labeled in an a priori manner as either intrinsic or extrinsic. Further research on the theory has continued along these lines. For example. Centers and Bugental (I966) classify self-expression, the interest value of the work, and feelings of satis- faction derived from the work itself as intrinsic sources of job satis- faction. Their findings indicate that white-collar workers value these sources of intrinsic motivation more than do blue-collsu: workers. The theoretical significance of such research remains unclear, however, as long as intrinsic motivation has no explanatory power and, indeed, such classifications are in themselves arbitrary. -3- Although the status of intrinsic motivation as a psychological. construct is unclear f we would argue that the concept is of considerable interest from still another perspective. Instead of asking what intrinsic motivation is and how it operates, it may be viewed as a perception on the part of individuals. That is, suppose that individuals attempt to label their behavior in motivational tsrms much as do motivational theorists. The seeds of such an approach have been developed by de Charms (1968) as part of his work on personal causation as an affective determinant of behavior, de Charms argues as follows t As a first approximation, we propose that whenever a person experiences himself to be the locus of causality for his own behavior (to be an Origin), he will consider himself to be intrinsically motivated. Conversely, when a person perceives the locus of causality for his behavior to be external to himself (that he is a Pawn), he will consider himself to be extrinsically motivated £1966, p. 328]. For de Charms, the crux of the distinction between intrinsic and extrinsic motivation stems from the feeling or perception of personal causation. Satisfaction derives from ar. activity which is perceived as intrinsically motivated because of our need to feel a sense of personal causation in our actions. de Charm's ideas may be readily extended to a more general approach to intrinsic motivation ty means of Bern's (l967a, 1967b, 1970, 1972) self- perception theory. According to this theory, a person infers his internal states by observing his own behavior and the context in which it occurs. Thus a person may label his behavior as intrinsically motivated under some conditions and as extrinsically motivated under others. The environment provides cues as to whether one's internal motivation is intrinsic or extrinsic. -4- One such cue, for instance, might well be the nature of the goal- object. Althoiigh there is as yet little direct evidence to support a self-perception approach to intrinsic motivation, an experiment by Kiesler and Sakumura (1966) is illustrative. Subjects received either $5*00 or $1.00 (goal object) for stating a position known to be in agree- ment with their own thoughts about an issue. Subsequently, the opinions of subjects in the $5.00 condition proved to be more susceptible to counter- communications. Evidently subjects in the $5»00 condition could be less sure that their behavior really reflected their true attitude. In terms of the present discussion, one interpretation of this finding might be that the greater the monetary payment, the greater the self-perception that one's behavior is extrinsically motivated rather than based upon intrinsic satisfaction. Taking a lead from Woodworth (19I8), a convenient way of viewing the self -perception process is to assume that an individual performs an intuitive meeins-encfe analysis of his behavior. As shown in Figure 1, Insert Figure 1 ?<.bout here different, self -perceptions may result according to the affect associated with the means and the ends of an action. Intrinsic motivation can be attributed most clearly when the means are positive and the ends are negative or neutral. Extrinsic motivation can be attributed when the means are negative or neutral eind the ends are positive. When both axe positive, the attribution may be iinstable. In this last case, an individual may seek to clarify the issue or simply assume that he is either intrinsically -5- or extrinsically motivated. Which of these he assximes may well depend on personality factors such as Rotter's (1966) dimension of internal versus external control or on situational noi.ns» The self-perception of intrinsic motivation may lead to a number of consequences. A person might, fcr e:cample, persist in the behavior even in the absence of a goal -object, simply because he has perceived his behavior as intrinsically rewarding. Likewise, actual performance of the activity may be enhanced through increased attention to the activity itself. Moreover, the self -perception of intrinsic motivation may directly produce satisfaction if it is correlated with other factors such as an increase in self-esteem, competence ('rfhite, 1959 )» or feelings of personal causation (de Charms, I968). de Chatrms' (I968) discussion of intrinsic motivation poses yet another interesting problem. Common sense would lead one to expect that intrinsic and extrinsic motivation summate to produce satisfaction, and most orgeinizational theories of job attitudes have made this assumption (e.g., Porter and Lawler, I968; Vroom, 196^). However, from a self- perception perspective, the combination of positive meauns and positive ends may be unstable. In fact, de Charms argues that intrinsic and extrinsic motivation may interact. Specifically, the introduction of extrinsic rewards for a behavior may decrease overall motivation rather than enhance it, because the rewards decrease the perception of intrinsic motivation. He also predicts, conversely, that motivation may be enhanced if a rewajcd is withheld, Fort\inately, there aire several recent studies bearing on this interaction between intrinsic and extrinsic motivation. Since these studies are oux mAJor source of evidence about a self-perception -6- account of intrinsic motivation, they need to be exajroined carefu3J.y» Empirical Research Several studies have explored the effects of extrinsic rewards on an indlvidiial's intrinsic motivation to perform a task, Deci, one of the more active researchers in this eirea, has followed de Charms in predicting that intrinsic and extrinsic rewards are not additive in their effect on motivation, ajid that the introduction of contingent monetary rewards or piuiishment reduces intrinsic motivation to perform an activity (Deci, I97I; Deci aj\d. Cascio, 1972). However, luilike de Charms, Deci has also predicted that verbal reinforcement increases intrinsic motivation to perform a task (Deci, 1971> 1972a), while non-contingent financial rewsirds leave Intrinsic motivation intact (Deci, 1972b), We will examine the support for each of these propositions and their relationship to a self -perception account of Intrinsic motivation. Table 1 provides a convenient summary of the studies reviewed. Insert Table 1 about here Contingent Rewards Research on the effects of contingent monetary rewards on intrinsic motivation is most relevant to the de Charms hypothesis. To explore these effects, Deci (l97l» 1972a) had subjects work on a series of interesting puzzles. The experimental sessions were divided into work periods and free-time periods, ]>iring a free-time period subjects could read magazines, -7- reraain idle, or continue to work on the puzzles? the araounb of free time . spent on the puzzles was taken on face validity as a dependent raeasxire of intrinsic motivation. The results of these studies indicated that subjects who were paid contingent on their task performance (the number of puzzles solved) spent less free time on the puzzles than did unpaid controls. Deci interprets this finding, as demonstrating that the contingently paid subjects lose intrinsic motivation for the activity. Their re-evaluation of the activity is produced by the perception that "it is motivated by the money" rather than "it is intrinsically motivated" (Deci, 1972a, p. 114). Although these contingent reward studies do seem to support a self- perception account, there are problems with their interpretation iriiich should be made explicit in order to direct further research. RLrst, in none of the Deci studies are the performance data reported for the experimental task. It is thus xinclear whether any chemge in free-time spent on the task is due to a change in perception or merely to a chauige in performance. That is, in terms of a causal model, the performstnce of the subjects is aui uncontrolled variable which possibly mediates the relationship between the manipu2.ated variable and the dependent measure of intrinsic motivation. On^j would expect, for example, that the intro- duction of contingent rewards increased effort in solving puzzles during the experimental sessions. Therefore, the decreased amount of free time spent on the puzzles after the experimental sessions could be due to factors such as satiation or fatigue rather than any cognitive re-evaluation of why one is performing the task. A second aimbiguity in the contingent reward studies is the magnitude -8- of the reward. Since the revards were administered contingent on performance, we have no information about the amoimt of reinforcement which actually constituted the manipulated variable. This omission is unfortunate since Deci (l972b) later compares the data from a contingent reward experiment (Deci, 1972a) to data fron a study using a fixed, noncontiogent reinforcement. In a comparison of intrinsic motivation resulting from these two different manipulations, it is thus impossible to tell how much the measure of intrinsic motivation differs because of differences in the contingency of the payment or in the amount of reinforcement delivered, A third ambiguity concerns the timing of the reward. In the contingent reward studies, extrinsic rewards decreased intrinsic motivation when subjects expected the reward out were not actually paid until after the completion of the entire experiment, including the free~time period. However, in one of these same experiments, Deci (l972a) reports data showing that intrinsic motivation increased when contingent payment was made after the task but before the free-time period. This latter finding was inter- preted as supporting equity theory si ce increased free-time spent on the task could have provided a means of resolving over-payment inequity. Whatever the merits of this interpretation, j.t still should be noted that the data provide a relevant, albeit nonconforming, test of a self -perception account. That is, one would expect payment before the free-time period to have made the extrinsic reward even more salient, thereby increasing the self -perception of extrinsic rather them intrinsic motivation. In short, there does not seem to be any obvious theoretical rationale for limiting the self -perception effect to contingent rewards presented after the free-time period. Equity considerations provide only a po st hoc -9- explanation for the absence of the effect when the reward vfas presented 3 before the free-tine period. Another issue relevant to the occurrence of rewards is vfhether they axe expected or not;. In the experiment just described, Deoi manipulated the timing of the reviard but subjects alvrays expected to receive the rewcird, Lepper, Greener and I'.'isbett (197&) vanned the expectation as well as the level of rewards. An extrinsic reward (a "Good Player Award" consisting of a gold star and red ribbon) was promised to soir.e chj.ldren before they performed an interesting tp,sk (playin^i with magic markers) while other children were not told of the award until aitor completion of the task. The extent to which the children played with th*^ magic markers was later recorded in a free play situation. The airiount of fcee- time spent on the markers was lowest for the group expecting tho reward and highest, for the group not expecting the revard, vrlth a no-reward control group interaGdiate between the two extremese Thus it is possible that any decrease in irttrinsic motivation is limited to expected extrinsic reiraxds • The Leppers Greane, and ilisbett lincirg suggests the possibility that it is not the self-parception of motivation that is the crucial factor but rather the parceptio.n of the offer of the revards Tha.t is, it is possible that the extri nsic reward j.s perceived as a bribe or as conveying information tliat the experimenter does: not /iew the acci\'lty as enjoyable enough to be performed without an extra reward, Knd.. as Steiner (l97l) points out, if an individual interprets a rev?ard as a bribe, he may perceive his personal freedom as being threatened, Ihus, it may be his reaction to this loss of perceived freedom which adversely affects his -10- subsequent performance (e.g., Brehm, I966) rather than a chaiige in intrinsic motivation. More research is needed on this possibility before we can be confident that the decline in performance is produced by a decrease in the self -perception of intrinsic motivation as opposed to a reaction to the offer of the reward. PunislTment , Verbal Reinforcement , and Noncontinsent Rewards Several other studies are also relevant to the interaction between intrinsic and extrinsic motivation r Deci and Gascio (1972) tested the hypothesis that punishment (a noxious buzzer) for poor task performance would decrease intrinsic motivation. The results showed that subjects who were threatened with contingent punishment spent (marginally significant) less free-time on a puzzle task than did controls. Again, consistent with a self -perception account, the introduction of an extrinsic force apparently reduced intrinsic motivation. However, unlike the previous studies of contingent rewards, it is also consistent with any alternative explanation based on stimulus generalization. The aversive stimulus, threat of punishment, inay have generaJ-ized to puzzle-solving durii^g the experimental session, with the negative association maintained for the same activity during the I'ree-time period. The possibility of this alternative explanation must be eliminated before any study of punishment can be taken as clear e\'idence for a self -perception account, Deci has also attempted to determine the effect of verbal reinforce- ment on intrinsic motivation. One experiment (Deci, 1971) manipulated verbal rewards using the same design employed for monetary rewards. Subjects again worked on a puzzle-solving task and then were exposed to a free-time period. Some subjects, but not others, were told that they -11- had performed much better than average or. the puzzle task. The results of this experiment showed that the verbally reinforced subjects spent more free-time on the puzzle than controls. A second experiment (Deci, 19?2a) combined the verla-l revrard variable idth the previously discussed manipulation of the timing of the contingent monetary reward (noney after the free-time period, money before the free-time period, no money). In this study, the effect for verbal reinforcement is nonsigxiificant by conventional standards. Nonetheless, Deci attempts to salvage the verbal reinforcement effect by interpreting a nonsignificant verbal reinforcement by sex interaction, and verbally reinforced male subjects do appear to si)end more time on the puzzles during the free time period. In any case, Deci (1972a, 1972b) interprets these two experiments as indicating that verbal reinforcement is not phenomenologiceilly distinct from intrinsic rewards and, therefore, adds to one's intrinsic motivation to perform a task. Certainly the evidence for this proposition is most airibiguous. Furthermore, as with pualsiiment, stimulus generalization or secondary reinforcement appear to be alternative interpretations to any self- perception account. Let us now turn to a study (Deci, 1972b) recently reported in this journal on the effects of noncontingent rewards on intrinsic motivation. As in earlier experiments, subjects participated in eui experimental work session and free-tine period. In this study, however, subjects in the experimental group vrere paid $2.00 regardless of their performance on the puzzle task, while control subjects were not paid at all. The results indicated no significant difference in the free-time spent on the puzzle between the experimental and control groups. This failure to find a -12- significant difference was interpreted as demonstrating that noncontingent monetary rewards do not change intrinsic motivation, auid that 'rfith non- contingent revards subjects aire less likely to perceive themselves as being motivated by the rewards. Thio conclusion is not justified by the data, Notice thax Deci has essentially affirmed the null hypothesis. Because we can never icnov? what factor, if any, accounts for a lack of change, it is logically impossible to prove tha absence of an effect. Was the receipt of noncontingent rewajrds in this experiment the sane as receiving no treatment at all, or v.-ere there other variables which caused subjects' intrinsic motivation to ramain intact? With respect to this saiae iioncontingent reward study, we should also note that Deci atteiapts to make his results more mearfngful by comparing them with two cells (money after-no verbal reinforcement and no money-no verlal reinforcement) from an earlier contingent monetciry rewards ex- periment (Deci, 19?2a), Unfort-onately there are severe problems in comparing two studies in this ways i\ny changes in the two experiments could have produced the different res-'ilts, Deci states in passing that the small, nonsignificant difference in the two no -reward (control) groups could be due "to the fact that a different experimenter conducted the two studies (i97^.b, p. 226)," but it is also clear tb^t the difference in the two reward groups could be due to x.he same kind of faxitors. In any event, the more appropriate interaction test is not reported, A recent experiment by Kruglanski, Rriedman, and Zeevi (1970) pro- vides better evidence about the effects of noncontingent rev.-ards on intrinsic motivation. In this study., some children, but not others, were offered an extrinsic reward (an interesting laboratory tour) for -13- participating in an experiment. The rev?ard, as in Deci's (l972b) ex- periment, was not contingent upon high parformaLnce but upon participation in the activity, Gontrari'- to Deci's hypothesis, a decrease in intrinsic motivation K-as obt.aincd for the extrinsic, noncontingent reward condition. Children who were offered the extrinsic revard were less satisfied tdth the experimental task aind less likely vmarginaily significant) to volvmteer for similar experimentse In addition, the extrinsically rewarded group did not perform as well on the experimental task (in terms of recall, creativity, and the Zeigarnik effect) as the nonrewarded childreno While more research is needed to determine whether contingent and noncontingent rewards both decrease intrinsic motivation, this study and the Lepper et al, study discussed earlier indicate that they do. In terms of a self -per- ception theory, if contingency maites rewards more salient as extrinsic forces, perhaps the effect is actually an interaction such that both con- tingent and noncontingent rewards decrease intrinsic motivation, but contingent re^Jards produce the largest chan£;e, 'flieoretical Implications What then is the status of the de Chrcrmn hypothesis' that intrinsic and extrinsic factors interact versus the tiaditional assumption of organization theory that they are 3,dditive? Ob'/iously the present ex- periraentaJ. evidence is inconclusive, though it does pro'/ida a basis for further research. The Deci (1971, 1972a) studies suggest that under some conditions the presence of an extrinsic reward rrtay reduce beh^-vior which we intuitively associate with being intrinsically motivated. In view of the Kruglanski et al , and Lepper et_ al , experiments, this effect does not appear to be llRiitGd to contingent extrinsic rewards a We STiggested rather that both contingent and noncontingent revfards decrease intrinsic motivation but, due to their .^jreater salience, contingent rewards may have a larger effect, (l/hetiacr verbal approval can piccke a task more intrinsically motivating is really bsside the point here.) However, even disregarding some of the luethodologioal problems noted, these re- sults only indicate that self -perception theory is relevant, not the form of such a theory. It may indeed be that extrinsic rewards lead individuals to perceive that they aro not intrinsically motivated aiid that this perception affects their subsequent performance. On the other hand, the Lepper et si , finding" that only expacted extrinsic rewards decrease intrinsic motivation raises the possibility that it is the perception of why the reirard is bein^ offered (i.e., a bribe vs. a lx>nus) vnilch affects performancce In short, the form of sslf -perception theory most relevant to these studies is by no means settled. In this research; there also needs to be more attention given to the dependent variables ?,ppropriace for study, Dacl and. Lepper et f.l» used persistence on a task as their single measures cf intrinsic motiva- tion. Nonptheless, -there ar« ooher indicators which can and should be utilized in atoessing intrintic motivation » Perhaps the most obvious indicator is reported task saoisfacticn- sinc^e one certainly should like a task if he is '.filling to perforrii it for no other apparent r3v?ard, Kruglanski et al, did report '-hat task enjoyment of the non-rewarded subjects exceeded tba.t of subjects who received an extrinsic reward, D'ecl reported that subjecxs rated the puszle task for interest and en- joyment at the end of the experimental sessions in his 1971 study. -15- Althoiigh Deci found that rated task satisfaction did not differ between the experimental amd control groups or among the experimental sessions, he does not riention the apparent inconsistency of this attitudlnatl data with his observed behavioral change. A much less direct indicator (or perhaps product) of intrinsic motivation would be any changes in actual performance on a task. The Deci and Lepper et al, experiments present data on task persistence and not on task performance. It is plausible that Intrinsic motivation may be related to organizational variables such as absenteeism and turnover, and the Kruglamski et^ a.l » data suggest a relationship between intrinsic motivation eind qualitative aspects of performsince. It should be stressed, however, that the linkage between task persistence and quantitative as- pects of performance is virtually unknown. Any implication that the process which makes people intrinsically motivated to perform an activity is the same process which causes people to perform well (in terms of quantity and quality) may contain the same pitfalls as the early hypotheses linking satisfaction and productivity (cf. Erayfieid and G?x)ckett, 1955)« As already noted, all of the dependent variables in the studies discussed were designed to measure Intrinsic motivation, de Chairras, however, originally stated his interaction hypothesis with overall task motivation as the dependent variable. Along these lines, an alternative research strategy is to follovr de Charms and treat both intrinsic and extrinsic factors as independent variables, measuring their effects on overall motivation. It seems to us that such a design might provide a fruitful test of the interaction hypothesis. Finally, one last point is particularly crucial for future research. -16- As pointed out earlier, there appaajr to be many problems confronting explanations of behavior based on the phenomena of intrinsic motivation. Although the concept seems descriptive, it is difficult to characterize as a psychological process. This is, in fact, why we argued that self- perception theory provides a valuable nev perspective on intrinsic motiva- tion. However, researchers have not aJLways maintained this distinction between intrinsic motivation as a perception sr.d as a phenomenon, Deci implicitly contends that extrinsic rewards decrease the perception of intrinsic motivation and this in turn decreases actual intrinsic motivation. We would axgue that a clear distinction should be made between the two. Too little is known about intrinsic motivation as a psychological process even to assume a direct relationship vdth the perception of intrinsic motivation. The attributed cause of a behavior need not be veridical with its objective cause* Tnus, research should attempt to relate the self-perception of intrinsic motivation to task performance and attitudes without making any premature assumptions about the actual existence or 14- nature of intrinsic motivation. Applicationn Based upon a review of his experiments ^ Deci has proposed that orgaxiizations should abandon contingent reward schemes and substitute noncontingent rewards in their place. The importance of the present noncontingent payment study is that money does not decrease intrinsic motivation if it is paid noncontingently. It is possible to pay workers and still have them intrinsically motivated. Hence, the writer favors the prescription that wo concentrate on structuring situations and jobs -17- to arouse intrinsic motivation, rather than trjrlng to structure piece-rate cuici other contingency pay- ment schemes, Workers vould be intririslcally moxlvated and T-rould seek t^ satisfy their hi£,her ordei' needs th^xugh effective performance rDeci, 1972b, p. 22?J, Interestingly enough t ordtitia^; data indicate thc^t organizations do not in fact use money as an incentive linked to perfonruuice (e«go> Halre, Ghlselli, and Gordon, 1967 ). Even so- in ',riew of our discussion of the limitations of Deci's findings on con-cingent versus noncontingent rewards, Deci's prescription for policy makers is definitely premature. Beyond the specif3.c merits of Deci's findings, however, it is important to realize that the results oi any study in this area should be applied with extreme caution. Any advice must be based not only on data regarding task persistence but also on multiple measures of task perfomance. Instead of more assunptions that the intrinsically motivated, happy wcrksi is a • productive worker ? we must have iriormation on the relationship between the perception of intrinsic motivation and performance under various task demands. It would appear plausible that intrinsic motivation is associated with the qualitative aspects of parformajice (cf, Kruglanski et al , , 1970) and that the incrinsically rotivated worker might provide surveillajxce over his own output (Katz, 196^1-) « However, the intrinsically motivated worker may not be the most productive in terns of his quantity of out- put or effort expenditiare (cf, Lawler, 1971), As a result, if research on a self -perception theory of motivation is to be applied constructively, we need both a theory and data on relevant task demands, A reasonable prediction is that the organization may reqiiire both the intrinsically motivated and the extrinsically motivated worker to function effectively. -18- tn order to have the Intrinsically motivated worker, though, attention oust be given to the worker's perception of organizational rewards as well as to the job itself. References Atkinson, J, W, An introduction to mo tivation « Princeton, N, J« j Van No strand, 196^+, Bern, D, J, Self-percepti on I The dependent variable of human performancea OrRajii zati o nal B ehavior and Kiman Perfor Birijice, 1967* 2, 105-121, (a) Bon, D, J, Self -perception! An alternative interpretation of cognitive dissonance phenomenao Psychological Review , 1967# 2!i* 183-200, (b) Bern, B, J, Beliefs , attitudes , and human affairs, Monterey, California! Brooks/Cole, 1970, Bern, D, J, Self -perception theory. In L, Berkowitz (Sd, ), Ad varices in experimental social psycho lo^ « Vol, 6, New York* Academic Press, 1972, Pp, I-S2I Bolles, R, C. Reinforcement, expectancy, and learning. Psychological Review , 1972, 79, 39^-^09. Brehm, J, M, A theory of psycholo,^cal reactance . New York? Academic Press, 1966, Centers, R, aind Bugental, D, Intrinsic sind extrinsic Job motivations among different segments of the working population. Journal of Applied Psychology . I966, ^, 193-197. Gofer, G» N, and Appley, H, H, Hotivatlon i Theory and research . New Yorki Wiley, 1967. de Charms, R. Personal causation t The internal affective determinants rf behavior, New York? Academic P. ass, 1968, Deci, E. L, The effects of externally mediated rewards on intrinsic motivation. Journ al of Pe rsonality and Social Psychology , 1971 » 18 , 105-115. Deci, E, L. Intrinsic motivation, extrinsic reinforcement, and Inequity, Journal of Personality and Social Psychology » 1972a, 22, 113-120, Deci, E. L, The effects of contingent and noncontingent rewards auid controls on intrinsic motivatio Performance, 1972b, 8, 217-229, controls on intrinsic motivation. O rgan! zational Behavior and Human Deci, E, L, and Cascio, W, F, Changes in intrinsic motivation as a function of negative feedback and threats. Paper presented at the Eastern Psychological Association Convention, April, 1972, Haire, M,, Ghiselli, E, E«, and Gordon, M, E, A psychological study of -21- pay. Journal of Appliod Fsychology Konocrraph , 196?, 51 » Herzberg, F. Work and the nature p ^ maii « Glev63.and, Ohiot World Pub- lishing Co., 196?7~ Hunt, J, HcVa Intrinsic motivation and its role in psychological develop- nent. Nebraska S:\rmpcsium on Motivation , 19^5 ^ 12» 189-282, Katz, D, The motivationrd basis of organizational behavJ-or, Behaviora-L Science , 1964, £, 131-146 « Kiesler, C« A. and Sakumura, J, A test of a model for commitment. Journal of Personality and Social Psychology , 1966, 2.» 3^9-353. Koch, S. Behavior as "intrinsically" regulatedt Work notes towards a pretheory of phenomena called notivational. Nebraska Symposium on Motivation , 1956, k, 42-8?. Kruglanskl, A. W,, Freedman, I,, and Zeevi, G. The effects of extrinsic incentives on some qualitative aspects of task performamce. Journal of Personality . 1971, 22* 606-6I7, Kruglanski, A. W, , Alon, S,, and Lewis, T. Retrospective misattribution and task enjoyment. Journal of Experimental Social Psychology , 1972, 8, 493-501. lawler, E. E. "Paj^ and o x-gani zational effectiveness ! a psychological view « New Yoricj McGraw-Hill, 1971. Lepper, M, Ra, Greeno, 1),. » and Nisbsttj R, E» Undermining children's intrinsic interest with extrinsic rewards s A test of the "over justifica- tion" hypothesis, j. of Personality and Social Psychology, in press. Porter, L« Mo and Lawler e S, E. Ma-n;u