Desire for Control and Achievement-Related Behaviors · Burger, Department of Psychology,...

Journal of Personality and Social Psychology1985, Vol. 48, No. 6, 1520-1533

Copyright 1985 by the American Psychological Association, Inc.0022-3514/85/S00.75

Desire for Control and Achievement-Related Behaviors

Jerry M. BurgerUniversity of Santa Clara

A model is presented that describes the relation between individual differences inthe general desire to control events and performance in achievement-related tasks.Six experiments were conducted with a college-student population to examinevarious steps in this model. Subjects high in the desire for control displayedhigher levels of aspiration, had higher expectancies for their performances, andwere able to set their expectancies in a more realistic manner than were subjectslow in the desire for control. Subjects high in desire for control were also foundto respond to a challenging task with more effort and to persist longer at adifficult task than were subjects low in desire for control. Finally, a pattern ofattributions for success and failure was uncovered for subjects high in desire forcontrol that has been associated with high achievement levels.

The construct of personal control has gen-erated a considerable amount of research re-cently (cf. Baum & Singer, 1980; Garber &Seligman, 1980; Lefcourt, 1981, 1982; Perl-muter & Monty, 1979). The prediction of,motivation for, and reaction to the loss ofpersonal control has been tied to numeroussocial-psychological phenomena. These includedepression (Abramson, Seligman, & Teasdale,1978), academic performance (Dweck & Licht,1980), health (Wallston & Wallston, 1981),pain tolerance (Thompson, 1981), crowding(Schmidt & Keating, 1979), and adjustmentamong the elderly (Schulz, 1980).

A natural extension of this research hasbeen the examination of individual differencesin the motivation to obtain control or to bein control of a situation. In this vein, Burgerand Cooper (1979) introduced the notion ofdesire for control, a stable personality traitreflecting the extent to which individualsgenerally are motivated to control the eventsin their lives. Persons high in desire forcontrol are said to prefer making their owndecisions, taking action to avoid a potentialloss of control, and assuming leadership roles

Portions of this research were presented at the meetingof the American Psychological Association, Anaheim,California, August 1983.

The author would like to thank Cynthia Allen, WilliamKraft, and Alicia Lopes for their help with the datacollection.

Requests for reprints should be sent to Jerry M.Burger, Department of Psychology, University of SantaClara, Santa Clara, California 95053.

in group settings. Persons low in desire forcontrol are motivated to avoid extra respon-sibilities and may prefer that someone elsemake decisions for them.

To examine this individual difference, Bur-ger and Cooper (1979) designed the Desir-ability of Control (DC) Scale. The DC scalehas been found to have reasonable psycho-metric properties, is only slightly correlatedwith measures of locus of control, and isrelatively independent of social desirability(Burger, 1984; Burger & Cooper, 1979; Smith,Wallston, Wallston, Forsberg, & King, 1984).Working with the scale, researchers havefound that DC scores can account for signif-icant proportions of variance in a wide varietyof areas. These include learned helplessness(Burger & Arkin, 1980), gambling (Burger &Cooper, 1979; Burger & Schnerring, 1982),choice of a place to die (Smith et al., 1984),intrinsic motivation (Burger, 1980), atti-tude change (Burger & Vartabedian, 1980),depression (Burger, 1984), perceptions ofcrowding (Burger, Oakman, & Bullard, 1983),and type of photographs taken (Henry &Solano, 1983).

The present series of experiments is de-signed to examine the role of individualdifferences in the desire for control inachievement-related behaviors. Research onachievement motivation and achievement-re-lated behavior extends from classic earlyworks (e.g., Atkinson, 1964; McClelland,1961) to recent interest in more complexmodels and in a variety of related variables

1520

DESIRE FOR CONTROL 1521

Theoretical

High DC as Comparedwith Low DC

High-DC Benefit

High-DC Liability

Aspiration Level

Select harder

more realistically

Higher goalsare achieved

May attemptgoals toodifficult

Heaponsa lo

Challenge

React with

Difficult tasks

May developper'ormance-inhibitingreactions

Persistence

Work at difficult

Difficult tasks

are completed

May invest toomuch effort

Atlrtbutlons for

SUCCMS A Failure

More likely to

to self and failure

Motivation levelremains high

May develop anillusion ofcontrol

Figure I. Four-step model for the relation between desire for control and performance on achievement-related behaviors.

(e.g., Spence & Helmreich, 1983; Thomas,1983). The present series of studies representsan initial effort to examine how individualshigh and low in desire for control differ intheir behavior at several steps along anachievement-behavior sequence.

I propose that individuals high in the desirefor control will display many of the behaviorsthat are related to higher achievement. Morespecifically, desire for control will have asignificant influence on this behavior at fourdifferent steps in the task sequence. The stepsin this sequence are illustrated in Figure 1.First, persons high in desire for control shouldhave higher levels of aspiration for their per-formances than do persons low in desire forcontrol. In addition, these individuals shouldselect aspiration levels that more realisticallyreflect their potentials. These predictions fol-low from the conception of an achievementtask as a challenge to the individual's percep-tion of personal control. Success at the taskdemonstrates one's ability to control suchsituations and one's mastery generally. Onthe other hand, failure to conquer the taskmay be interpreted as a threat to one's per-ceived ability to control significant segmentsof the environment. Thus, persons high inthe desire for control should be highly moti-vated to perform well on a challenging task,and this should be manifested in high levelsof aspirations. However, unrealistically highaspirations can result in the unwanted failureexperience. Therefore, persons high in desirefor control should have learned to realisticallyset their aspirations in a manner that maxi-mizes their perception of achievement.

The second step in the model concerns the

individual's response to a challenge. Manyreal-world tasks in achievement settings arefilled with unexpected difficulties. I hypothe-size that persons high in desire for controlwill respond to these difficulties as challengesto their control over the task. Thus, thereactance effect (Brehm, 1966) of exertinggreater effort when one meets such a challengeshould be found more often among thosehigh, rather than low, in the desire for control.

Similarly, persons high in desire for controlshould be more persistent in their efforts tocomplete a difficult task—the third step inthe sequence. Because they are highly moti-vated to avoid failure, and thus to avoid theperception of a lack of control, persons highin the desire for control should work at adifficult task for a longer time before givingup or seeking help than do persons low inthe desire for control.

Finally, individuals high in desire for con-trol should attribute the causes of their per-formances in a manner that increases moti-vation on subsequent tasks. Research by Wei-ner et al. (1971) demonstrated that highachievement is associated with giving oneselfcredit for successes (e.g., high ability, higheffort) and attributing failure to a lack ofeffort or to luck. These attributions are saidto influence how the person approaches andworks on future tasks; this pattern leads tohigher motivation and thus more achieve-ment. I propose that persons high in desirefor control are motivated to perceive them-selves as responsible for their successes andthereby satisfy their motives to feel personallyin control of the situation. In addition, a highdesire for control should lead these same

1522 JERRY M. BURGER

individuals to deny any inability to controlthe situation (i.e., failure) and, therefore, toattribute such instances to unstable sources.

Although it is intuitively appealing on thesurface, this four-step model would be overlysimplistic in describing the relation betweendesire for control and performance onachievement-type tasks. Although the hy-potheses all point to the superiority of theperson high in the desire for control over theindividual low in desire for control, this iscertainly not always the case. At each step ofthe model it is possible that a high desire forcontrol also can become a liability. A highlevel of aspiration, for example, may lead tothe person's attempting tasks that are toodifficult to be accomplished. The person highin desire for control also might be more likelyto take on too many tasks, thus inhibitinghis or her ability to perform well on any ofthem.

Similarly, although persons high in desirefor control may react to challenges with in-creased effort, at some point the inabilityto control the situation may cause themto develop performance-inhibiting reactions.Wortman and Brehm (1975) proposed thatthose individuals who react most strongly toa threat to perceived control are most likelyto suffer helplessness reactions if the situationremains uncontrollable. Consistent with theWortman and Brehm model, Burger andArkin (1980) found that persons high in thedesire for control exhibited more helplessnessbehavior and reported higher levels of depres-sion following a learned helplessness manip-ulation than did subjects low in desire forcontrol. Similarly, Burger (1984) found thatpersons high in the desire for control whogenerally perceived that they did not havecontrol over the events in their lives weremore likely to exhibit some depressive symp-toms than were persons low in desire forcontrol. Thus, although the tendency to reactto a challenge with greater effort may bedesirable for some tasks, it also may holdsome serious pitfalls.

Whereas persistence on a task can oftenlead to the solving of a difficult problem, italso can lead to an inefficient investment oftime and effort if the task eventually provesto be too difficult. Finally, although attribu-tions to oneself for success may lead to higher

levels of motivation on later tasks, this ten-dency also can become a liability. Not rec-ognizing when one's successes are caused byexternal forces and not accepting responsibil-ity for one's failures may give the individualhigh in desire for control a false impressionof his or her abilities. Persons high in thedesire for control mistakenly may attempt tocontrol events over which they have little orno influence. Burger and Cooper (1979) andBurger and Schnerring (1982) found thatpersons high in desire for control were moresusceptible to the illusion of control thanwere persons low in the desire for control.Under certain gambling situations these sub-jects tended to place larger bets, indicative ofa greater confidence of success, for gamesthat were obviously chance determined thandid subjects low in desire for control.

The complete model for the relation be-tween desire for control and achievement-related behavior is presented in Figure 1. Ascan be seen, a complete understanding of thisrelation requires that the model be tested atnumerous points to determine the circum-stances under which different levels of desirefor control lead to different types of behaviorsand different outcomes. The purpose of thepresent series of experiments is to test theproposed positive relation between desire forcontrol and behaviors associated with highlevels of achievement at each of the foursteps in the model. Limiting conditions andthe influence of other variables on theserelations need to be examined in later re-search.

Experiment 1

1 propose that desire for control is positivelyrelated to level of aspiration on an achieve-ment-type task. Persons high in the desire forcontrol should be more likely to select tasksthat provide a challenge than should personslow in desire for control. In this way thecompletion of the task can provide a senseof mastery and control. Although the selectionof a relatively easy task might ensure success,such an accomplishment would provide littlesatisfaction for the person who is motivatedto perceive himself or herself as in control.On the other hand, persons low in the desirefor control are not as strongly motivated to


sense control through achievements, andtherefore may be satisfied to accomplish thenecessary work with a minimal chance offailure. These persons low in desire for controlmay be expected to select a task that allowsfor completion, even though it may not pro-vide much in the way of perceived control. Itwas predicted that when given a choice oftasks of different difficulty, persons high inthe desire for control would tend to selecttasks that are more difficult than would per-sons low in the desire for control.

Method

Subjects. Thirty-four male and female undergraduatesserved as subjects in exchange for class credit. All hadtaken the Desirability of Control Scale (Burger & Cooper,1979) approximately 4 weeks earlier as part of a largetest battery. No connection was made between the exper-iment and the DC scale at the time of the experiment, aprocedure that was followed in Studies 2-5, as well.

Instrument. The DC scale is a 20-item inventory thatasks subjects to indicate on 7-point scales the extent towhich they agree or disagree with statements concernedwith issues of control (e.g., "I prefer a job where 1 havea lot of control over what I do and when I do it," "I wish1 could push many of life's decisions off on someoneelse"). Scores range from 20 to 140, with higher scoresindicating a higher desire for control. Means with collegestudent samples have consistently averaged between 100and 105.

Procedure. Subjects participated in the experimentin groups. They were told that the experimenter wasinterested in examining verbal abilities. Subjects wereinformed that they would be working on a series ofanagram tasks. It was explained that the presentation ofthe anagrams would be in two parts. During the firstpart each subject received a test booklet containing foursets of 10 anagrams, one set per page. The experimenterexplained what anagrams were and gave an example.Subjects then were given 2 min each to solve the foursets of 10 anagrams, with the experimenter starting andstopping subjects for each set. All anagrams consisted offour letters and were designed to be only mildly difficult.

Subjects then were instructed to turn to the last pageof the booklet. On this page, subjects found the instruc-tions for selecting the level of difficulty for the anagramsthey would work on during the second part of theexperiment. It was explained that the subject would workon three more sets of 10 anagrams. The subject was toindicate which three of the possible nine sets he or shewished to work on and in what order they were to begiven. It was explained that the sets had been normedwith a large number of college students and that each setwas identified on the sheet with a difficulty rating. Thedegree of difficulty for each set was indicated by thepercentage of subjects in the normed group who suppos-edly were unable to solve all 10 anagrams. The nine setsranged from anagrams in which 90% of the normedgroup had been unable to solve all of the problems to a

set in which only 10% had been unable to complete allof the anagrams. The remaining sets were presented in10% increments. Thus, each of the three choices madeby the subjects could be assigned a value from 1 to 9 forthe degree of perceived difficulty. Subjects were told theycould select each of the sets only once.

Following these choices, subjects were informed thatthere would be no more anagram problems. The subjectswere debriefed and dismissed.

Results and Discussion

Subjects were divided into high- and low-desire-for-control (DC) halves via a median-split method. First, each group was comparedfor performance on the anagram task. Thetotal number of solved anagrams on the foursets (40 total anagrams) was calculated foreach subject. As can be seen in Table 1, nosignificant difference was found between high-and low-DC subjects in terms of performanceon the anagram task.

Next, subjects' aspirations were assessed intwo ways. First, the anagram set that thesubject desired to work on first was examined.Next, the three choices of anagram set weresummed for an overall aspiration score. Ascan be seen in Table 1, significant differencesbetween high- and low-DC subjects emergedon both of these measures. High-DC subjectschose an anagram set for their first choicethat was significantly more difficult than theset chosen by low-DC subjects, F(\, 32) =5.34, p < .03. In addition, the overall aspi-ration score for high-DC subjects was higher—indicating a choice of more difficult ana-grams—than was the score for the low-DCsubjects, F(l, 32) = 5.30, p < .03.

The results of Experiment I, therefore,provide support for the first step in the desire-for-control-achievement-behavior modelproposed here. Although high- and low-DCsubjects performed equally well on the sampleanagrams, the high-DC subjects preferred towork on tasks that they believed to be ofgreater difficulty than did the low-DC subjects.In actual achievement situations, high-DCpersons can be expected to attempt jobs orto take classes that are perceived to be difficultmore readily than will low-DC persons. Thisis consistent with the conception that thehigh-DC individual finds satisfaction in dis-playing mastery over tasks that he or shefinds challenging.


Table 1Mean Correct Anagrams and Choices by High-and Low-DC Subjects

Subjects

Variable High-DC Low-DC

Number correctFirst choiceTotal choice

35.306.12

19.35

35.414.29

16.23

Note, DC = desire for control.

Experiment 2

If desire for control is related to level ofaspiration in an achievement-related task,then one would expect that higher DC scoreswould be related to higher expectancies ofsuccess on the task. In addition, it has beenproposed that higher levels of desire for con-trol will be associated with more realisticexpectancies and aspirations. High-DC peopleshould not aspire to levels too high to meetand thus exclude a feeling of accomplishment.In addition, because high-DC individuals aremotivated to perform well, they may paycloser attention to performance feedback andadjust their aspirations accordingly more thando low-DC persons. Experiment 2 comparedexpectation levels of high- and low-DC sub-jects. It was predicted that high-DC subjectswould give higher and more accurate estimatesfor their performances than would low-DCpersons.

Method

Subjects. Eighty-five male and female undergraduatesserved as subjects in exchange for class credit. All hadtaken the DC scale several weeks earlier as part of a largetest battery.

Procedure. The procedure was similar to that usedin an experiment by Snow (1978). It was explained tosubjects that they would be working on a series ofconnect-the-numbers puzzles to assess their perceptualand motor skills. Subjects received a booklet containingthe puzzles. It was explained to subjects that each puzzleconsisted of the numbers 1-50, scattered randomly aroundthe 8Vi X ll-in. (21.8 X 28-cm) page. Their job was tobegin with the number 1 and connect the numberssequentially as rapidly as possible. Subjects were informedthat they had 20 s to work on each puzzle.

On the first page of the booklet, subjects were askedto indicate how many numbers they thought they wouldbe able to connect within the 20-s lime period for thefirst puzzle. The experimenter then started and stoppedsubjects on each of the six puzzles in the booklet. Aftercompleting each puzzle, subjects were instructed to in-

dicate on the spaces provided at the bottom of the pagehow many numbers they had connected on that puzzleand to estimate how many they would connect on thenext puzzle.


Three different measures were created fromthe subjects' responses. Each of these wasthen correlated with the DC score. First,there was a significant correlation betweenDC score and the estimate subjects made forthe first puzzle (r = .31, p < .01). The higherthe DC score, the higher the subject tendedto estimate his or her performance on theunseen task. Next, the six estimates for thepuzzles were summed for an overall aspirationmeasure. Once again, this score was positivelycorrelated with the DC score (r = .28, p <.01), indicating higher overall aspiration levelsfor those with higher DC scores. Finally, ameasure of subject accuracy of estimationwas calculated. For each puzzle the completednumber was subtracted from the estimate forthat puzzle. The absolute values of each ofthese figures were then summed for an ac-curacy index. This accuracy score correlatednegatively with the DC score (r = -.38, p <.001). Thus, higher DC scores were associatedwith smaller differences between the estimatefor a puzzle and the actual performance onthe puzzle.

The results of Experiment 2 provide ad-ditional support for the first step in theproposed model. On both the initial puzzleand the puzzles taken together, high-DC sub-jects tended to make higher estimates of theirperformance than did low-DC subjects. Theseestimates can be interpreted as another dem-onstration of the higher levels of aspirationproposed for high-DC persons. In addition,high-DC persons tended to be more accuratein their estimates of their performances. Thisfinding is consistent with the description ofthe high-DC person as one who is highlymotivated to demonstrate his or her mastery.These subjects set their aspirations on a levelwhere they were more likely than were low-DC subjects to attain the goal and, thus, toestablish their competence.

Experiment 3

The second stage of the model proposesthat persons high in the desire for control


will respond to a challenging task with greatereffort than will low-DC individuals. A taskthat becomes more difficult than was origi-nally anticipated should be perceived by thehigh-DC individuals as a threat to their con-trol over the task. These individuals shouldthen respond with greater effort to overcomethe challenge than should low-DC persons.In research on a related construct, Fazio,Cooper, Dayson, and Johnson (1981) foundthat Type A college students worked harderon a proofreading task when the task wasmade challenging than when the task was arelatively simple one. Type B persons showedthe opposite effect. Glass (1977) suggestedthat the primary distinction between Type Aand Type B persons is a higher level ofmotivation to control the environment by theType A individual. Indeed, Musante, Mac-Dougall, Dembroski, and Van Horn (1983)found that DC scale scores were significantlycorrelated with various measures of Type A-Type B. Because of this proposed similaritybetween the Type A-Type B construct andthe DC construct, both were examined in theexperiment.

A procedure similar to that used by Fazioet al. (1981) was employed. Subjects workedon a proofreading task that was either rela-tively easy or fairly challenging. It was pre-dicted that higher desire-for-control scoreswould be associated with higher levels ofeffort on the proofreading task, but only inthe challenging condition.

Method

Subjects. Thirty-nine undergraduates participated inthe experiment in exchange for class credit. All had takenthe DC scale and a measure of Type A-Type B, thestudent version of the Jenkins Activity Survey (JAS;Glass, 1977), several weeks earlier.

Procedure. Subjects participated in the experimentin groups. Subjects were given a red marking pen and amanuscript filled with errors. They were informed thatthey would be working on a proofreading task. Theexperimenter explained that the subject's task was tolook for errors in spelling, punctuation, and grammar.Examples of how to indicate and correct errors weregiven. In the challenge condition, subjects were given twoadditional tasks to perform. They were instructed tocount the number of times the word the appeared andalso the number of proper nouns that appeared on thelines they were proofreading. Subjects in the no-challengecondition were given the proofreading task only.

All of the subjects were given 10 min to work on thetask. At the end of the time, the experimenter stopped

the subjects and instructed them to draw a line indicatingthe last full line of the manuscript that had been proofread.Subjects in the challenge condition were instructed toindicate how many times they had read the word the andhow many proper nouns they had seen to that point inthe manuscript. All subjects then were debriefed.


Subject scores for the DC scale and theJAS were significantly correlated (r = .39,p < .02). Because of this correlation, analysisof the dependent measure, the number oflines attempted, was conducted with multipleregression analyses. The DC and JAS scores,along with a challenge-no-challenge dummyvariable, were entered into the regressionanalysis, producing an overall R2 of .18. Asignificant effect for the challenge variablewas found when it was entered into theanalysis first, F(l, 35) = 4.17, p < .04. Sub-jects in the no-challenge condition attemptedmore lines (M - 35.1) than did subjects inthe challenge condition (M =30.1). TheDC X Challenge interaction fell short ofsignificance, F(l, 35) = 2.39, p < .13, andthe AB X Challenge interaction was far fromsignificant (F < I ) . No other significant effectswere found in these analyses.

To better understand the relation betweenDC score and the dependent measure, sepa-rate regression analyses were conducted forthe challenge and no-challenge conditions.First, for subjects in the challenge condition,the JAS score was entered into the regressionanalysis. The results indicate that the JASscore was not a significant predictor of thenumber of lines attempted (F < 1). With thevariance accounted for by the JAS removed,the DC score was entered into the regressionanalysis. The DC score proved to be a signif-icant predictor of the measure, F(l, 16) =4.25, p < .05. When entered into the regres-sion analysis in the opposite order, the DCscore, entered first, approached significance,F(\, 16) = 3.69, p < .07, whereas the JASscore, with the DC variance removed, againfailed to predict the measure. The higher theDC score, the more lines the subject in thechallenge condition was likely to have at-tempted. When the total lines measure wasexamined within hierarchical multiple re-gression analyses for the no-challenge subjects,no significant effects were found.


The results of Experiment 3, therefore,provide some support for the second step inthe desire-for-control-achievement-behaviormodel. It was found that higher desire forcontrol was associated with greater effort onthe proofreading task, as indicated by thenumber of lines attempted, when the taskwas made a challenging one by the additionof two tasks. When the task was a simpleproofreading task, no effect for desire forcontrol was found. Although the experimentwas a fairly vague laboratory task, the behav-ior in the challenge condition can be seen assomewhat typical of behavior in real worksettings. At higher levels of achievement,workers often are called on to perform tasksthat require the ability to alternate one'sfocus of attention and to coordinate severaldifferent aspects of a task. The results suggestthat persons high in the desire for controlreact to such challenging jobs with greatereffort in an attempt to overcome the challengeand establish the perception of control overthe situation. However, no differences betweenhigh- and low-DC persons may be expectedfor jobs that present relatively low levels ofchallenge. Although they probably are moti-vated to avoid failure in this situation, high-DC individuals are not challenged enough toput forth the increased effort, as they are onthe more difficult task.

Although the above analysis is consistentwith the proposed model, note that subjectsin the challenge condition had a more difficulttask and therefore attempted fewer proof-reading lines overall than did the no-challengesubjects. Because of this, it is not possible todetermine if the high-DC subjects reacted tothe challenge with increased effort or if thelow-DC subjects responded to the challengewith decreased effort, or if both conditionswere the case. Nonetheless, the difference inhigh-DC and low-DC persons' reactions tothis situation has been demonstrated.

One additional interesting finding was thefailure of the Type A-Type B variable topredict behavior in either of the two types oftasks. Although there are some methodolog-ical differences between the two experiments,Experiment 3 represents a failure to concep-tually replicate the Fazio et al. (1981) findings.It is also interesting to note that althoughthey are conceptually similar, the desire for

control and Type A-Type B constructs weredemonstrated to have a practical and empir-ical distinction.

Experiment 4

The third step in the proposed model isconcerned with the individual's persistenceon a difficult task. I propose that personshigh in desire for control are more likely topersist at working on a difficult task than arepersons low in the desire for control. For ahigh-DC individual the difficult task may beperceived as a challenge to his or her abilityto control the situation. To give up on sucha task is tantamount to admitting that onehas encountered a task that he or she cannotcontrol. It obviously would be incorrect toassert that high-DC persons never admit thatthey can not accomplish a task. However, ina situation in which it is strongly impliedthat the task is not impossible, merely diffi-cult, the high-DC person can be expected tocontinue efforts to complete the task muchlonger than can the low-DC individual, whomore readily admits to an inability to controla task. The high-DC worker who is given adifficult task on the job naturally will assumethat the task can be completed with enougheffort, and thus can be expected to producethat effort for a long period of time.

Subjects in Experiment 4 were presentedwith some problems for which, in reality,there were no solutions. Although they weregiven the option of moving on to anotherproblem when they found the current prob-lem too difficult, high-DC persons, it waspredicted, would work at the unsolvableproblems longer before giving up and, there-fore, work at fewer problems during a speci-fied period of time than would low-DC indi-viduals. The amount of time spent workingon an impossible task has been used in thepast as a measure of persistence, and thispersistence has been tied to higher levels ofachievement motivation (Feather, 1961).

Method

Subjects. Tuenty-five male and female undergraduatesserved as subjects in exchange for class credit. All hadtaken the DC scale approximately 4 weeks earlier as apart of a large test battery.

Procedure. Subjects participated in the experimentin groups. The experimenter explained that the study


was concerned with testing verbal skills. It was explainedto subjects that they would be working on a series ofCryptoquote puzzles. The experimenter then passed outa sheet explaining what Cryptoquote puzzles were and,through examples, gave hints on how to solve them. Theexperimenter read the instructions aloud as the subjectsread along. It was explained to subjects that the puzzlesconsisted of a series of letters that represented a well-known quote or phrase. The object of the task is tofigure out the code and thereby discover the answer tothe puzzle. The task is to discover which letters the lettersin the puzzle actually represent. For example, an A mightbe represented by an F, such that every time an Fappeared in the puzzle the subject would know that itwas really an A.

After the experimenter determined that all of thesubjects understood the task, a booklet containing thefirst three Cryptoquote puzzles was distributed. To aidthe subject, a few of the correct letters were already filledin on these puzzles. Subjects were given 7 min each towork on the three puzzles. At the end of the 7 min, theexperimenter stopped the subjects, gave the correct answer,but told subjects not to write the correct answer in thebooklet if he or she had not already found the solution.

After the subjects completed the third puzzle, theexperimenter distributed a second booklet containingseven puzzles, one per page. Subjects were told that theywould have a total of 15 min to work on this set ofpuzzles. It was explained to subjects that they were freeto spend as much or as little time as they desired on anyof the seven puzzles. The only restrictions were thatsubjects were to place a check mark in the appropriatespace on the puzzle sheet before beginning each puzzleand that once subjects had turned past a puzzle in thebooklet, they were not to return to that puzzle. Theserestrictions were included to allow the experimenter todetermine if subjects had worked on the puzzles even ifno marks appeared on the page.

There were no letters already filled in on the puzzlesto help the subjects in the second booklet. In reality,none of the puzzles was solvable. The absence of helpwith the second set of puzzles made the subjects' transitionfrom solving the first three anagrams to total failure onthe second set credible. No subjects expressed any sus-picion that the puzzles might not be solvable. After 15min, the experimenter stopped the subjects and debriefedthem fully.


Subjects were divided via a median splitinto high- and low-desire-for-control halves.These two groups then were compared ontwo different measures. First, the number ofCryptoquote puzzles successfully completedfrom among the first three puzzles was deter-mined for each subject. As can be seen inTable 2, nearly all of the subjects did quitewell on this part of the task and did notdiffer by desire-for-control level. Next, thenumber of puzzles attempted during the 15

Table 2

Mean Performance and Persistence onCryptoquote Puzzles

Subjects

Variable

No. of puzzles solvedNo. of impossible puzzles

attempted

High-DC

2.92

3.31

Low-DC

2.67

4.50

Note. DC = desire for control.

min session was determined for each subject.A lower number of attempted puzzles wasindicative of greater persistence on the im-possible task. As shown in Table 2, the twoDC groups differed significantly on this mea-sure, F(l, 23) = 6.85, p < .02: The high-DCsubjects worked on significantly fewer puzzlesthan did the low-DC subjects.

The results thus provide support for thethird step in the proposed model. High-DCsubjects were found to work longer at difficult(in reality, impossible) tasks than were low-DC subjects. Consistent with the model, itcan be said that these subjects were stronglymotivated to deny that they were unable tocontrol the task. Note that there was a strongimplication that the tasks were indeed solv-able, especially after the subject had per-formed relatively well on the earlier puzzleswith the aid of the letters provided by theexperimenter. Naturally, once the high-DCindividual concludes that the task is an im-possible one, he or she probably is no morelikely than is the low-DC person to continuefruitless efforts.

Experiment 5

The final step in the model proposed herefor the relation between desire for controland performance on achievement-related be-haviors is concerned with attributions forsuccessful and unsuccessful performances. Ithas been found that people who are high inachievement motivation and who generallyreach high levels of achievement tend toexhibit a general pattern of attributions forsuccess and failure (Weiner et al., 1971). Thispattern, attributing successes to oneself andfailures to unstable sources, is said to lead to


higher levels of motivation on subsequenttasks. The reason for this motivation followslogically. If I believe a success was caused bymy own abilities or efforts, I am likely tocontinue my efforts on subsequent tasks inorder to continue my success. On the otherhand, if I believe that the success was causedby something outside of my control, then Iam unlikely to perceive that my efforts willassure success in the future. If I attribute myfailure to something stable, such as my lackof ability or a task that will always be thisdifficult, then there is little chance of successin the future and little reason to make astrong effort on subsequent tasks. On theother hand, if I believe that my failure wascaused by a lack of effort on this particulartask, than a strong effort on the next taskshould result in success.

Subjects in Experiment 5 were exposed tosuccess and failure experiences. It was ex-pected that subjects high in the desire forcontrol would be more likely to attributesuccessful outcomes to their own abilitiesthan would low-DC subjects. This is becausehigh-DC individuals should be motivated toperceive that their successful efforts are rep-resentative of their ability to control thesituation. What is unclear is whether high-DC persons are more or less motivated to seetheir successes as internally caused whentheir overall performance has been poor. Onthe one hand, it is possible that these subjectsare motivated to see the entire task perfor-mance as being controlled by chance, therebyallowing for the belief that the overall failureexperience does not represent the person'sgeneral ability level. On the other hand, one'sperceived control over the task may be threat-ened with an overall failure experience, thusmore strongly motivating the high-DC personto attribute his or her few successes to himselfor herself. In Experiment 5, therefore, theextent to which subjects attributed their suc-cessful outcomes to their own ability versusluck was assessed in overall success and failuresituations. It was expected that high-DC sub-jects would attribute more of their perfor-mance to ability when they were successfulthan would low-DC subjects. How DC levelwould affect the interpretation of the perfor-mance was less clear in the overall failuresituation.

Table 3Number of Correct Guesses Attributed to Skill

Condition

Subjects Failure Average Success

High DCLow DC

5.27

4.62

8.577.00

20.14

12.20

Kate. DC = desire for control.

Method

Subjects. Sixty male and female undergraduates servedas subjects in exchange for class credit. All had taken theDC scale several weeks earlier as part of a large battery

of tests.Procedure. A procedure similar to that used by

Weiner and Kukla (1970, Experiment 5) was employed.Subjects were told that the experimenter would read a

list of 50 numbers, all either / or 0, in a particular order.The subject's task was to guess the number before the

experimenter announced it on each of the 50 trials.Subjects were told that the numbers were not presentedin a random order, but rather that there were generaltrends in the number presentation testing the subject'sanalytic ability. In reality, the numbers were presented in

a random order.After completing the 50 trials, subjects were told to

indicate the number of correct guesses and then tocomplete a short questionnaire. Two key items on the

questionnaire were taken from the Weiner and Kukla(1970) experiment. These were to indicate how many ofthe correct guesses were the result of skill rather than

lucky guessing, and to estimate how many correct guessesthey would make if asked to perform the task again witha new set of numbers. Following completion of thequestionnaire, subjects were fully debriefed.


Subjects were trichotomized into failure,average, and success groups according to theirnumber of correct guesses. Subjects also weredivided into high- and low-DC groups via amedian-split procedure. A 2 X 3 (High-DC-Low-DC X Failure-Average-Success Perfor-mance) analysis of variance was conductedon the two dependent measures. Two signifi-cant main effects emerged on the skill versusluck measure: desire-for-control effect, F(l,54) = 4.12, p < .05, and performance effect,F(2, 54) = 7.25, p < .01. The interaction wasnot significant. As shown in Table 3, high-DC subjects attributed their performance toability more often than did low-DC subjects.In addition, the better subjects did on the


task, the more they tended to credit theirabilities. Planned comparisons found thathigh- and low-DC subjects differed signifi-cantly in the success group (p < .05). butnot in the other two performance groups.

Two significant main effects also emergedon the measure asking subjects to estimatetheir performance on a second guessing task:desire-for-control effect, F(l, 54) = 4.26, p <.05, and performance effect, F(2, 54) = 5.86,p < .05. Once again, the interaction was notsignificant. Means for these effects are pre-sented in Table 4. As seen in the table, high-DC subjects predicted that they would dobetter than did low-DC subjects on a secondtask in all three performance conditions.Planned comparisons between high- and low-DC subjects failed to reach significance inany one of these conditions, however. Withregard to the performance main effect, sub-jects in the success condition exhibited thegreatest confidence on an upcoming task,whereas subjects in the failure condition ex-hibited the least amount of confidence.

Experiment 5 therefore provides supportfor the fourth step in the proposed model.High-DC subjects were more likely to attrib-ute their successful performance to skill (ascompared to luck) than were low-DC subjects.This attribution pattern has been shown tobe related to high levels of achievement. Notethat the skill-versus-luck item was selectedbecause it was the most logical for the partic-ular task employed. It seems that subjects inthis type of laboratory experiment all ap-proach the short task with comparable levelsof effort. Indeed, the task was described asone in which the key determinants of one'sperformance were one's ability and luck.Naturally, attributions have been found to befar more complicated than this. But for thepurposes of the present series of experiments,the relation between desire for control andattributions posited by the model has beendemonstrated. It should also be noted thathigh-DC subjects were more confident thatthey could perform well on a subsequenttask, regardless of condition. This is consistentwith the findings of Experiment 2, whichdemonstrated the high-DC individual's higherexpectancy as reflected in a higher level ofaspiration than that held by the low-DCperson.

Table 4

Number of Correct Guesses Estimated forAnother Task

Condition

Subjects Failure Average Success

High DC 26.67Low DC 25.50

33.7127.57

36.4234.40

Note. DC = desire for control.

Experiment 6

A final experiment was conducted to ex-plore the relation between desire for controland attributions. It has been found that in-dividual differences in desire for control arefairly stable personality dimensions. It can beinterpreted from this that the pattern ofattributions postulated by the model anddemonstrated in Experiment 5 also shouldbe fairly stable across time and situations.One way to examine this possibility and toprovide further support for the model is toassess and compare desire for control with ameasure of stable attribution patterns.

Peterson et al. (1982) developed the Attri-bution Style Questionnaire (ASQ) to assessstable patterns of attributions. The ASQ de-scribes 12 situations, half with positive out-comes and half with negative outcomes. Sub-jects are asked to give a major cause for whyeach of these outcomes would happen tothem. They are then asked to indicate on aseries of 7-point scales the extent to whichthis cause is internal, stable, global (appliesto many situations), and important. Thecombined totals for each of these items pro-vides a measure of the subject's general ten-dency to make attributions along each ofthese four dimensions.

It can be predicted that desire for control,when applied to the desire-for-controlachievement model, would correlate positivelywith the tendency to make internal attribu-tions, but only for situations with positiveoutcomes. Desire for control should correlatenegatively with the tendency to give internalattributions for situations with negative out-comes. Similarly, desire for control should bepositively correlated with the tendency to givestable attributions for positive events, but


negatively correlated with the tendency togive stable attributions for negative events.Finally, it can also be predicted that desirefor control will be positively correlated withthe individual's tendency to see positive out-comes as caused by factors that will apply tomany situations (global), but be negativelycorrelated with the tendency to give globalattributions for events with negative outcomes.There is no clear prediction for the relationbetween desire for control and the tendencyto see success or failure as important. It ispossible that desire for control is positivelyrelated to both of these, for the high-DCindividual might interpret both success andfailure as being more important than mightlow-DC persons because both of these, successand failure, may reflect on the individual'sability to control the situation. This, however,is very speculative at this point.

Method

Subjects. Sixty-one male and female undergraduatesserved as subjects in exchange for class credit.

Procedure. All subjects were administered the DCscale and the Attribution Style Questionnaire. Half ofthe subjects received the DC scale first, and half receivedthe ASQ firsl.


Subjects' responses to the ASQ were dividedinto those referring to events with positiveoutcomes and those referring to events withnegative outcomes. For each of these sub-scales, a score for internality, stability, glob-ality, and importance was calculated. Eachof these eight scores was then correlated withthe DC score. The correlations are presentedin Table 5.

As can be seen in Table 5, there was someweak but consistent support for the hy-potheses. Desire for control was positivelycorrelated with the tendency to make internal,stable, and global attributions for events withpositive outcomes, although these correlationsare not strong and fall below the acceptedlevel of significance. There was also a veryweak finding for desire for control to benegatively correlated with the tendency toattribute events with negative outcomes tointernal, stable, and global factors. Thesecorrelations are so small, however, that they

Table 5Correlations Between DC Scaleand ASQ Subscales

Events

Attribute Positive Negative

InternalitvStabilityGlobalityImportance

.13

.22*

.22*

.18

-.11-.02-.08

.02

Note. DC = Desirability of Control; ASQ = AttributionalStyle Questionnaire.* p < .10.

can not be interpreted as supportive of thehypotheses.

The small size of the correlations can betraced to several possible sources. First, thereliabilities of the subscales of the ASQ aredisturbingly low; Peterson et al. (1982) re-ported internal reliability coefficients for thesescales ranging from .44 to .69. Second, thescale contains some items that deal withaffiliation situations and not achievement sit-uations per se. This may weaken the appli-cability of the ASQ to the achievement model.Given these weaknesses in the scale, thecorrelations for the events with positive out-comes presented in Table 5, in combinationwith the results of Experiment 5, can beviewed as providing some additional supportfor the desire-for-control-achievement-behavior model.

General Discussion

The six experiments provide support forthe model outlined earlier concerning therelation between individual differences in ageneral desire to control events and perfor-mance on achievement-related behaviors.Some support was found for each of the foursteps in the model. Subjects high in the desirefor control were found to aspire to higherlevels of achievement than subjects low indesire for control. These high-DC personsalso were found to have higher estimates fortheir performances and were able to adjusttheir expectations in a more realistic mannerthan were low-DC individuals. Subjects highin desire for control were found to respondto a challenge with more effort and to persist


at a difficult task longer than were low-DCsubjects. Finally, high-DC subjects exhibitedan attribution pattern for successes and fail-ures that suggests a high level of motivationon subsequent achievement tasks.

Although the results of the experiments,using a variety of laboratory tasks, consistentlysupport the model, there are several pointsof caution that need to be addressed briefly.First, as with any model, more experimentalsupport needs to be obtained. The model andresearch reported here are intended as aninitial step in describing the relation betweendesire for control and achievement behavior.It is very likely that the model will be ex-panded and modified as more research isaccumulated.

Second, the importance of examining thehypotheses in real-world settings with differentsubject populations needs to be emphasized.It is impossible to create a laboratory situationthat includes all of the complex forces thatinfluence behavior in actual achievement set-tings. Such variables as competition fromboth inside and outside the organization,economic conditions, personal goals and dis-tractions, supervision, economic incentives,affiliation among workers, and time effectscan influence the manner in which desire forcontrol and achievement behavior are related.For example, Burger (1980) found that high-DC subjects were especially susceptible to theundermining of intrinsic motivation with theintroduction of extrinsic rewards. It is possiblethat working for money in a real job andworking for the undefined rewards availablein the laboratory task may be perceived quitedifferently by people high in the desire forcontrol. In addition, the present series ofexperiments may be limited by a restrictedrange of DC scores. It may be expected thata wider range of individual differences indesire for control will be found in a moreheterogeneous population than the studentsample used here. This could account forsome of the low correlations found in theinvestigations.

Finally, more elaboration is needed tospecify when a high desire for control facili-tates performance and when it becomes aliability. As outlined earlier, at each step inthe model it is possible that an individualhigh in the desire for control will engage in

behaviors that ultimately are unsuccessful orinefficient. A better understanding of the in-teraction between the type of task and thelevel of desire for control is needed if themodel is to increase in its predictive value.

Another question that emerges from thisresearch concerns the relation between thedesire-for-control concept and the traditionalneed for achievement, or achievement moti-vation, construct. At least three plausiblerelations can be suggested. First, it is possiblethat achievement motivation is the superor-dinate category, with the desire for controlbeing but one element that contributes to anoverall high level of motivation in achieve-ment settings. This view is consistent withthe approach outlined recently by Spenceand Helmreich (1983), in which it is arguedthat the need for achievement may not be aunidimensional construct. Spence and Helm-reich identified three components of achieve-ment motivation, some parts of which resem-ble the desire-for-control construct: the desireto work hard, mastery, and competitiveness.These investigators also have demonstratedthat the components do not relate to achieve-ment behaviors in exactly the same way. Ifdesire for control is but one component ofachievement motivation, then the present se-ries of investigations can serve to further ourunderstanding of how this one aspect ofachievement motivation is related to variousachievement-type behaviors.

A second possibility is that there are indi-vidual differences in a general desire to controlevents that can influence behavior in a largenumber of areas, of which traditionalachievement contexts are but one. Indeed,researchers have uncovered significant effectswhen examining a wide variety of non-achievement behaviors with the DC scale.From this perspective one can conceive ofachievement motivation as but one compo-nent of a more general desire for control.The third possibility is that neither the desire-for-control construct nor the achievement-motivation construct is superordinate, butrather each may be influenced by the other.A high desire for control, for example, maybe manifested in a high-achievement motive,and a high need for achievement may beenhanced by a strong desire to control events.Obviously, pinpointing the relation between


the two constructs provides substance forfurther investigation.

If further research proves to be successfulin detailing and refining the various stages inthe model, there appear to be some importantpractical implications. The most obvious ap-plication is the matching of worker and taskto maximize performance. Although it cer-tainly is not advisable to assign personnel onthe basis of one personality inventory, em-ployers may want to consider desire for controlalong with other personality and aptitudevariables when selecting persons for jobs orassigning jobs to employees. It may also bepossible to identify problem behaviors, suchas persisting too long on a difficult task ormaking external attributions for successes,and modify these through the application ofthe model. For example, high-DC workersmight be instructed to spend no more thana certain length of time on a particular taskand be assured that the inability to solve theproblem is not an indication of one's lack ofcontrol. Sensitive supervisors also might seeto it that these individuals have a large numberof success experiences to keep motivationlevels high. These possibilities are, of course,highly speculative at this time and awaitadditional investigation.

References

Abramson, L. Y., Seiigman, M. E. P., & Teasdalc, J. D.(1978). Learned helplessness in humans: Critique andreformulation. Journal of Abnormal Psychology, 87.49-74.

Atkinson, J. W. (1964). An introduction to motivation.Princeton, NJ: Van Nostrand.

Baum, A.. & Singer, J. E. (1980). Advances in environ-mental psychology: Application.1; of personal control.Hillsdale, NJ: Erlbaum.

Brehm, J. W. (1966). A theory of psychological reactance.New York: Academic Press.

Burger, J. M. (1980). E/feclance motivation and theoveriuslificatiim ejject. Unpublished doctoral disserta-tion, University of Missouri-Columbia.

Burger, J. M. (1984). Desire for control, locus of control,and proneness to depression. Journal of Personality,52.71-89.

Burger, J. M., & A r k i n , R. M. (1980). Prediction, controland learned helplessness. Journal ol Personality andSocial Psychology. JX. 482-491.

Burger, J. M., & Cooper, H. M. (1979). The desirabilityof control. Motivation and Emotion, 3, 381-393.

Burger, J. M., Oakman, J. A., & Bullard, N. O. (1983).Desire for control and the perception of crowding.Personality and Social Psychology Bulletin. 9, 475-479.

Burger, J. M., & Schnerring, D. A. (1982). The effects ofdesire for control and extrinsic rewards on the illusionof control and gambling. Motivation and Emotion, 6,329-335.

Burger, J. M., & Vartabedian, R. A. (1980). Desire forcontrol and reaction to proattitudinal and counterat-titudinal arguments. Motivation and Emotion, 4, 239-246.

Dweck, C. S., & Licht, B. G. (1980). Learned helplessnessand intellectual achievement. In J. Garber & M. E. P.Seiigman (Eds.), Human helplessness: Theory and ap-plications (pp. 197-221). New York: Academic Press.

Fazio, R. H., Cooper, M., Dayson, K., & Johnson, M.(1981). Control and the coronary-prone behavior pat-tern: Responses to multiple situational demands. Per-sonality and Social Psychology Bulletin, 7, 97-102.

Feather, N. T. (1961). The relationship of persistence ata task to expectation of success and achievement-related motives. Journal of Abnormal and Social Psy-chology, 63, 552-561.

Garber, J., & Seiigman, M. E. P. (1980). Human help-lessness: Theory and applications. New York: AcademicPress.

Glass, D. C. (1977). Behavior patterns, stress, andcoronarydisease. Hillsdale, NJ: Erlbaum.

Henry, W. P., & Solano, C. H. (1983). Photographic styleand personality: Developing a coding system for pho-tographs. Journal of Psychology, 115, 79-87.

Lefcourt, H. M. (1981). Research with the locus of controlconstruct (Vol. 1). New York: Academic Press.

Lefcourt, H. M. (1982). Locus of control: Current trendsin theory and research (2nd ed.). Hillsdale, NJ: Erlbaum.

McClelland, D. C. (1961). The achieving society. Prince-ton, NJ: Van Nostrand.

Musante, L., MacDougall, J. M., Dembroski, T. M., &Van Horn, A. E. (1983). Component analysis of theType A coronary-prone behavior pattern in male andfemale college students. Journal of Personality andSocial Psychology, 45, 1104-1117.

Perlmuter, L. C., & Monty, R. A. (1979). Choice andperceived control. Hillsdale, NJ: Erlbaum.

Peterson, C., Semmel, A., von Baeyer, C., Abramson,L. Y, Metalsky, G. I., & Seiigman, M. E. P. (1982).The Attributional Style Questionnaire. Cognitive Ther-apy and Research, 6, 287-300.

Schmidt, D. E., & Keating. J. P. (1979). Human crowdingand personal control: An integration of the research.Pswhologtcal Bulletin. g6. 680-700.

Schulz, R. (1980). Aging and control. In J. Garber &M. E. P. Seiigman (Eds.), Human helplessness: Theoryand applications (pp. 261-277). New York: AcademicPress, 1980.

Smith, R. A., Wallston, B. S.. Wallston, K. A.. Forsberg.P. R., & King. J. E. (1984). Measuring desire forcontrol of health care processes. Journal of Personalityand Social Psychology, 47, 415-426.

Snow, B. (1978). Level of aspiration in coronary proneand noncoronary prone adults. Personality and SocialPsychology Bulletin, 4, 416-419.


Spence, J. T., & Helmreich, R. L. (1983). Achievement-related motives and behaviors. In J. T. Spence (Ed.),Achievement and achievement motives: Psychologicaland sociological approaches (pp. 7-74). San Francisco:Freeman.

Thomas, E. A. C. (1983). Notes on effort and achievement-

oriented behavior. Psychological Review, 90, 1-20.Thompson, S. C. (1981). Will it hurt less if 1 can control

it? A complex answer to a simple question. PsychologicalBulletin, 90, 89-101.

Wallston, K. A., & Wallston, B. S. (1981). Health locusof control scales. In H. M. Lefcourt (Ed.), Research

with the locus of control construct (Vol. I, pp. 189-243). New York: Academic Press.

Weiner, B., Frieze, I., Kukla, A., Reed, L., Rest, S., &

Rosenbaum, R. M. (1971). Perceiving the causes ofsuccess and failure. Morristown, NJ: General LearningPress.

Weiner, B., & Kukla, A. (1970). An attributional analysis

of achievement motivation. Journal of Personality andSocial Psychology, IS, 1-20.

Wortman, C. B., & Brehm, J. W. (1975). Responses touncontrollable outcomes: An integration of reactancetheory and the learned helplessness model. In L.Berkowitz (Ed.), Advances in experimental social psy-chology, (Vol. 8, pp. 277-336). New York: AcademicPress.

Received May 27, 1983

Revision received January 31, 1984 •

Special Call for Papers on the Personality and Social Psychology of Aging

The Psychology arid Aging journal is gearing up for its first year of publication in 1986.

Manuscripts have been received in the editorial office for a number of months, but

more than 50% of the original submissions have been experimental. The Editor,

M. Powell Lawton, and the Associate Editor, Donald H. Kausler, wish to emphasize

that Psychology and Aging will be a broad-ranging publication, and manuscripts from

all areas of psychology are desired.

Papers on personality and social issues related to aging are encouraged. As the

proposed editorial policy statement outlines:

Psychology and Aging publishes original articles on adult development and aging. Such originalarticles include reports of research, which may be applied, biobehavioral, clinical, educational,experimental (laboratory, field, or naturalistic studies), methodological, or psychosocial. Whilethe emphasis is on original research investigations, occasional theoretical analyses of researchissues, practical clinical problems, or policy may appear, as well as critical reviews of a contentarea in adult development and aging. Clinical case studies that have theoretical significanceare also appropriate. Brief reports are acceptable with the author's agreement not to submit afull report to another journal; a 75-100 word abstract plus 48-space lines of text and referencesconstitute absolute limitations on space for such brief reports.

Manuscripts should be directed to:

M. Powell Lawton

Philadelphia Geriatric Center

5301 Old York Road

Philadelphia, Pennsylvania 19141

Desire for Control and Achievement-Related Behaviors · Burger, Department of Psychology,...

Documents

Transcript of Desire for Control and Achievement-Related Behaviors · Burger, Department of Psychology,...