Journal of Experimental Social Psychology Journal of Experimental Social Psychology 38 (2002) 368–383 www.academicpress.com

Priming against your will: How accessible alternatives affect goal pursuit James Y. Shaha,* and Arie W. Kruglanskib a

Department of Psychology, University of Wisconsin–Madison, 1202 West Johnson Street, Madison, WI 53706, USA b University of Maryland, College Park, MD, USA Received 16 October 2000; revised 26 March 2001

Abstract Four studies examined how diverse aspects of goal pursuit are influenced by the accessibility of alternative goals. It was consistently found that such an accessibility often affects the resources allocated to a focal goal, influencing commitment, progress, and the development of effective means, as well as one’s emotional reponses to positive and negative feedback about one’s striving efforts. Moreover, the direction of these influences was found to depend on how the alternative goals relate to the focal pursuit. Alternatives unrelated to the focal goal pull resources away from it, whereas alternatives facilitatively related to a focal goal draw resources toward it. Ó 2002 Elsevier Science (USA). All rights reserved.

When Mary enters her school library, her voice drops to a whisper and her thoughts turn to a paper due in her psychology class at the end of the semester. John’s current concerns always become quite social when he returns to his fraternity house after dinner. Such examples illustrate recent findings by Bargh and his colleagues, suggesting that goals may often be primed by the environmental context in which they are pursued (Bargh, 1990; Bargh & Gollwitzer, 1994; Bargh & Barndollar, 1996). Chartrand and Bargh (1996), for example, have demonstrated that goals can be unconsciously activated by semantically similar terms. Thus, words like ‘‘evaluate’’ may unconsciously prime impression-formation goals and the books in a library may bring to mind one’s goal to engage in studious contemplation. But goals are rarely primed in a motivational vacuum and may often be activated when one is consciously pursuing something else. Thus, while John’s fraternity house may often prime in him social intentions, the implications of such priming may depend on its relation to John’s presently chosen pursuit. How, then, might such social goal priming affect John when he has the explicit intention of doing something quite dif-

*

Corresponding author. E-mail address: jshah@facstaff.wisc.edu (J.Y. Shah).

ferent, such as studying? Perhaps such a conscious intention would simply shield John from any unconscious priming effects, making him impervious to the social temptations embodied by his fraternity house. On the other hand, given the extent to which goal priming has been shown to prompt goal pursuit and magnify its emotional significance (see Higgins, Shah, & Friedman, 1997), it is also possible that such priming may come to interfere with the conscious pursuit of a focal goal. The present work seeks to address these possibilities by examining whether goal priming affects one’s conscious focus on a current pursuit and, if so, to explore the nature of this effect as well as its implications for behavior and emotional experience. In various forms, the general notion that a goal pursuit may be affected by an alternative has long been entertained in lay and scientific notions of motivation. The 14th-Century French philosopher Jean Buridan, for instance, asserted that awareness of two equally desirable pursuits prevents progress on either. The extremity of his position was satirized by the parable of ‘‘Buridan’s Ass’’: a hungry ass stands between two equally appetizing bales of hay; although obviously motivated, the ass is unable to choose between them and eventually dies (see Turner, 1903). Yet classic research on motivation has offered at least a modicum of support for Buridan’s general perspective

0022-1031/02/$ - see front matter Ó 2002 Elsevier Science (USA). All rights reserved. PII: S 0 0 2 2 - 1 0 3 1 ( 0 2 ) 0 0 0 0 5 - 7

J.Y. Shah, A.W. Kruglanski / Journal of Experimental Social Psychology 38 (2002) 368–383

in acknowledging that the psychological presence of alternative goals may create an ‘‘approach–approach’’ conflict (Lewin, 1935, 1951; Miller, 1944; Zeigarnik, 1938). Such work illustrates the difficulty one may face in focusing on a given goal when appealing alternative goals exercise their ‘‘pull’’ on motivational resources. The recognition of such detrimental effects led Shallice (1972) to assume that individuals typically inhibit competing ‘‘action systems’’ whose simultaneous activation would drain the available resources. Similarly, Kuhl (1984) has argued that progress toward goal attainment involves both the active pursuit of a chosen focal goal and the inhibition of alternative goals that might come to mind. Consistent with the above analyses, research on selfcontrol has shown that our ability to ignore some goals has significant implications for our ability to attain others. Thus, Kuhl and Weiss (1985) found that for individuals focused on their internal states, failure to reach one goal may result in subsequent performance deficits in the pursuit of unrelated goals because of an inability to cognitively disengage from the initial failure (see also, Brunstein & Olbrich, 1985; Kuhl & Kazen-Saad, 1988). Moreover, when Kuhl and Helle (1986) interrupted depressed patients before they could complete a goal of cleaning up a messy desk, this interfered with a subsequent test of short-term memory, suggesting that patients suffered from an inability to cognitively ‘‘put aside’’ interrupted goal pursuits (see also Kuhl & Beckman, 1994). Mischel’s delay of gratification work (Mischel & Ebbeson, 1970; Mischel, Ebbeson, & Zeiss, 1972; Mischel, Shoda, & Rodriguez, 1989) similarly found that children are much better at keeping to their goal of waiting for the experimenter and ignoring the temptation to fulfill another goal instead (e.g., eating a cookie) when the competing goal-object (the cookie) is hidden from view than when it is clearly visible. Correspondingly, children who refrained from looking at the tempting object were better able to stick to their goal of waiting than children who chose not to look away. Though the foregoing findings are suggestive and consistent with the notion that the presence of alternative goals may interfere with current pursuits, the underlying process that governs these phenomena and its precise consequences for self-regulation have not been fully understood, leaving a number of fundamental questions unexplored. What exactly defines, for instance, the ‘‘presence’’ of an alternative goal and how consciously aware of such an alternative must one be for it to undermine one’s present pursuit? Moreover, what is the precise mechanism whereby alternative goals may undermine each other? Is this undermining effect simply due to the mental distraction posed by the accessibility of any other cognitive stimulus, or is it restricted to the specific motivational pull that only goals may exercise? Finally, what specific aspects of goal pursuit may be affected by the accessibility of alternative goals and what

369

are some of the possible boundary conditions on such an effect? While the notion of pull has traditionally suggested a ‘‘pulling away’’ of resources, thus undermining attainment, might there be occasions in which the presence of goal-alternatives draws resources toward a more concerted goal pursuit? That is, could there be circumstances wherein an accessible alternative actually serves as a reminder of one’s current pursuit, augmenting goal focus and the intensity of goal pursuit? The present research attempts to provide some preliminary answers to such questions by adopting a perspective on goal-directed strivings that emphasizes the cognitive aspects of motivational phenomena.

The cognitive side of motivation In social psychology, motivational and cognitive perspectives have often been viewed as rival interpretations of the same findings (cf. Bem, 1967; Bem, 1972; Kunda, 1990; Miller & Ross, 1975). Parallel to this research tradition, however, is one that has implicitly acknowledged the cognitive aspect of motivation. The early research on intrinsic motivation, for instance (cf. Deci, 1971; Kruglanski, 1975; Kruglanski, Friedman, & Zeevi, 1971; Lepper, Greene, & Nisbett, 1973) assumed that motivation may be inferred from one’s performance of an activity and the circumstances in which it has occurred. More recently, discrepancies from desired states have been viewed as cognitive structures whose degree of activation may vary chronically and as a function of the situation (Higgins, 1989, 1997) and, as discussed earlier, a similar perspective has been fruitfully applied to goals (Bargh & Gollwitzer, 1994; Kruglanski, 1996a, 1996b; Trzebinski, 1989). The present research shares the cognitive approach to motivation exemplified by the above work and applies it toward gaining a deeper understanding of the impact that the presence of alternative goals may exert on ongoing strivings. Specifically, we interpret this impact in terms of alternatives’ cognitive accessibility while pursuing a given focal goal: to the extent that alternative goals are accessible and unrelated to the focal goal, they may interfere with the pursuit of that goal by pulling away from its limited regulatory resources. Thus, we are assuming that increased accessibility of alternative goals, effected, say, by environmental priming (Bargh, 1990), may automatically pull away the attentional and motivational resources allocated to a focal goal when these alternatives are unrelated to the focal pursuit. In turn, the lowered resources devoted to the focal goal should also dampen one’s emotional reactivity to perceived progress or a lack of progress vis-a-vis that goal. If the presence of competing alternatives causes the focal goal to be pursued less ardently, then actors should be less anxious or dejected by feedback about a lack of progress but also be

370

J.Y. Shah, A.W. Kruglanski / Journal of Experimental Social Psychology 38 (2002) 368–383

less satisfied or relieved by feedback about progress. In addition, the pulling away of resources from the focal goal by its rival alternatives might reduce the individual’s ability to develop efficient means to goal attainment. As a consequence, one’s progress toward goal attainment might be hampered. In short, we are assuming that the mere cognitive activation of alternative goals unrelated to a current pursuit will foster a whole range of cognitive and motivational effects that will detrimentally affect attainment of the focal objective.

Facilitative inter-goal relations and the direction of ‘‘goal pull’’ effects Though an alternative goal may often pull attentional and motivational resources away from a focal pursuit, there are occasions wherein the attainment of one goal facilitates an attainment of the other. For instance, ‘‘becoming a successful professional’’ and ‘‘being liked and respected by others’’ may be seen as mutually facilitative objectives in that successful professionals are often liked; conversely, being liked and respected by others may be of help in attaining one’s professional objectives. The presence of such facilitative relations may reduce or reverse the effects posed by the accessibility of an alternative by drawing resources toward the focal goal. That is, while an unrelated alternative goal may pull resources away from a focal pursuit, a facilitatively related alternative may instead draw attention toward focal goal attainment because a facilitatively related goal may be regarded as a ‘‘means’’ for the focal goal’s attainment. A means, in turn, may prime its correspondent goal, which may bring to mind its inherent value or desirability and mobilize resources toward its pursuit. Shah and Kruglanski (2000) have recently obtained evidence that goals may be primed by their attainment means, defined as those activities that facilitate goal attainment. This priming, in turn, was found to increase goal focus and to intensify goal pursuit, even when the specific priming means was not utilized. This work suggests therefore that a focal goal may be primed by a facilitatively related alternative and that its increased accessibility due to such priming may benefit goal pursuit and attainment. To summarize then, we have assumed that (a) goals function as cognitive entities that may be made more accessible through environmental priming (Bargh & Barndollar, 1996); (b) the accessibility of alternative goals may automatically affect commitment to the focal goal which, in turn, may influence the formulation of effective means, the likelihood of goal attainment, and the emotional significance of goal pursuit. Finally, (c) the direction of these effects is determined by the extent to which the focal and alternative goals are perceived to have a facilitating relation to each other.

We performed four studies to explore these notions. These manipulated (through environmental priming) the simultaneous accessibility of an alternative goal with a focal task goal that participants were currently pursuing. We assumed that accessibility of unrelated alternative goals exerts a ‘‘goal pull’’ that automatically lessens commitment to a focal goal and reduces the development and selection of effective attainment means. Such an accessibility should also hinder attainment of the focal goal and dampen emotional responses to success and failure-feedback en route to its pursuit. Finally, we assumed that the undermining effects of such ‘‘goal pull’’ would be reversed when the alternative goals were seen as facilitatively related to the focal goal.

Study 1 Our first study sought to manipulate the accessibility of an alternative goal and examine the consequences this would have for attainment of a current task goal. The accessibility manipulation was accomplished through a subliminal priming procedure that repeatedly displayed the name of an alternative task goal, as participants were pursuing their primary or focal goal. We predicted that the increased accessibility of this alternative (through priming) would affect participants’ commitment to the focal goal, as reflected by their performance and persistence at the task, and that the nature of this effect would depend on the relation of the alternative goal to the focal pursuit. Method Participants Fifty-five students of the University of WisconsinMadison participated in the study in exchange for extra credit toward a course requirement. The sex of participants was not recorded in this particular study, though it was recorded in all our remaining experiments. Procedure Participants completed a majority of the procedure on IBM-compatible computers. Instructions to all participants stated that they would be completing two separate tasks, the first of which would involve the solving of anagrams. This initial task was described as a measure of ‘‘verbal recall,’’ portrayed as a valuable skill for writing. Each anagram consisted of a string of letters participants attempted to ‘‘unscramble’’ to find as many different words as possible using all of the letters. Participants were also told that individuals with good verbal recall could find 80% of the possible words in the set of anagrams. To introduce the alternative goal with which the participants would be subsequently primed, participants were informed that

J.Y. Shah, A.W. Kruglanski / Journal of Experimental Social Psychology 38 (2002) 368–383

after completing the anagram task they would be performing a different task to assess their skill at ‘‘functional thinking.’’ This task, referred to as the ‘‘box-use’’ task, would involve the naming of as many different functions for a box as they could think of in a set (and undisclosed) time period. Participants were told that they would be completing this second task immediately after they had completed the first one. No direct information about the interrelation of the two tasks was provided, although we assumed that participants might find reason to assume some relation from the fact that both involved the generation of as many different solutions as possible on each trial. But because this only hinted at a possible relation, we assumed that participants in our sample would naturally vary in the extent to which they perceived the two tasks as interrelated. Participants were then asked, ‘‘how difficult does the anagram task seem?’’ and ‘‘how related is verbal recall to functional thinking?’’ and provided answers on a 10point scale ranging from 1 (not at all) to 10 (extremely). The latter question intended to tap participants’ perception of a facilitative relation between the two task goals, assuming that if they seemed related, attaining one would indicate that one is likely to attain the other as well. As we had assumed, participants’ ratings of the tasks’ relatedness naturally varied across the entire range of the scale from 1 to 10, with a median score of 6. Prior to embarking on the actual anagram task and ostensibly to familiarize them with the procedure, participants performed five sample anagrams. It is at that time that the priming of the alternative task goal took place: specifically, before each sample anagram was presented, participants were subliminally primed with a word phrase. In the control condition, they were presented with the phrase ‘‘view it’’ for 50 ms that was backward masked with a non-word letter string (‘‘xxxxxxxxxxxxxx’’). In the experimental, ‘‘high accessibility’’ condition, they were presented with the phrase ‘‘box use’’ for 50 ms backward masked in the same manner. Participants then completed a set of 5 anagrams, comprising the actual anagram task. The computer recorded the answers they provided for each anagram. After completing the anagrams, participants indicated how important it was for them to do well on the boxuse task, how much they had enjoyed the anagram task, and how difficult they found it. Each response was given on the same 10-point scale described earlier. Finally, a funneled debriefing procedure was used to assess whether participants had guessed the true nature of the study and had seen the presented primes. Each participant was asked a series of questions concerning his or her experiences while completing the program. Participants were debriefed about the nature of the experiment and the fact that they had been presented with prime words. At this point, participants were

371

asked whether they had seen these primes and, if so, whether they could identify them. No participant reported any suspicion about the study before the debriefing nor did any report seeing the word primes after having been told that they had been presented on the screen. Anagram performance and persistence Participants’ performance on the anagram task was calculated by simply summing all the correct solutions that they found across the trials. Participants’ persistence on the anagram trials was calculated by summing their persistence on all the actual task trials after employing a natural logarithmic transformation on each individual trial time to lessen the impact of outliers and to correct for skewness in the distribution of persistence times (Judd & McClelland, 1989). Results Relation among variables An initial correlational analysis found that participants’ rating of the facilitative relation between the two tasks was not significantly related to how difficult or enjoyable they found the anagram task, r ¼ :09, p > :50, r ¼ .11, p > :40, respectively, but was marginally significantly related to how committed they were to the box-use task, r ¼ . 24, p > :10. This correlational analysis also found participants’ persistence on the anagrams to be significantly positively correlated with their performance, r ¼ :53, p < :001. Participants’ rating of the facilitative relation between the two tasks was then subjected to a median split to form separate high and low relation groups. An initial set of analyses examined whether the accessibility manipulation affected how difficult they found the anagram task, how enjoyable they found it, and how important they perceived the box-use task to be. Consistent with our expectations, accessibility had no significant effect on these ratings, F s < 1. Separate ANOVAs were then performed to examine the effect of accessibility manipulation on participants’ anagram performance and persistence for individuals who perceived the goals to be facilitatively related and for those who perceived them to be unrelated, controlling for their other ratings of the task. With all of these variables included in the analyses, the accessibility manipulation as such was not found to significantly affect how well participants performed and how long they persisted at the task F ð1; 48Þ ¼ 2:05, p > :15; F ð1; 48Þ ¼ 1:26, p > :25, respectively. Additionally, perceived goal interrelation as such did not have a significant effect on anagram performance or persistence, F s < 1, but it significantly interacted with accessibility to affect both dependent measures, F ð1; 48Þ ¼ 4:14, p < :05;

372

J.Y. Shah, A.W. Kruglanski / Journal of Experimental Social Psychology 38 (2002) 368–383

Fig. 1. Anagram performance and persistence as a function of alternative priming and perceived relation (Study 1).

F ð1; 48Þ ¼ 7:44, p < :01, respectively. According to this interaction, the detrimental effect of alternatives’ accessibility on anagram performance and persistence was significantly greater when the perceived facilitative relation between the goals was low than when it was high. Participants’ total anagram performance and untransformed persistence time (in seconds) as functions of the type of prime and perceived inter-goal relation are presented in Fig. 1. The results of Study 1 confirm our predictions. Accessibility of the alternative goal interacted in the predicted manner with its perceived relation to the focal goal in its impact on commitment to the focal goal, as assessed by participants’ task persistence and performance: accessibility undermined persistence and performance to a greater extent when the two task goals were perceived as unrelated than when they were perceived as facilitatively related. Note that these results cannot be explained by the effects of our accessibility manipulation on perceptions of the alternative goal in terms of its difficulty and importance ratings, which did not differ significantly across the two accessibility conditions. Nor can these results be explained by the mere effects of distraction due to a priming manipulation per se, in so far as such a manipulation occurred in our control condition as well. A more compelling explanation of these findings is that it was the accessibility of alternative goals that affected the resources allocated to the focal goal, depending on its perceived relation to the alternative. Our subsequent experiments attempted to conceptually replicate these findings while also testing additional aspects of our theory. Specifically, we investigated whether accessibility of an alternative goal and the perception of facilitative relations between it and the focal goal have the predicted joint effects on the development of effective attainment means with respect to the focal goal as well as on participants’ emotional reactivity to progress (or the lack of progress) toward that goal.

Study 2 Our analysis suggests that the accessibility of an alternative goal may affect the extent to which resources are pulled away from, versus being mobilized toward, pursuit of the focal goal. Among other things, this may affect the degree to which an individual would possess sufficient resources to invest in the independent development of appropriate attainment means with respect to that goal. When resources are pulled away, participants may often rely on ‘‘conventional’’ attainment strategies, or uncritically use means proposed by an external agent, instead of developing new means. Participants may be more likely to develop new means, however, when the alternative re-directs their attention back to the focal goal. Thus, the effect of an accessible alternative on the use of means to pursue a focal goal should also be moderated by the degree to which the alternative and the focal goals are perceived as facilitatively interrelated. Our present study examined the foregoing implications of our analysis and attempted to replicate the performance and persistence findings of Study 1. In most relevant respects, the experimental paradigm of Study 2 resembled that of Study 1. The focal goal again involved the solving of anagrams. Also, we again used a subliminal priming procedure to manipulate the accessibility of an alternative goal. Unlike Study 1, however, participants in the present study were also presented with a possible attainment means for pursuing the task goal. Pre-testing had suggested that the use of this particular means did not significantly increase the anagram performance. Nevertheless, because accessibility of an unrelated alternative goal was hypothesized to impair one’s ability to generate effective attainment strategies, participants primed with an alternative goal may more readily depend on the presented task means when solving the anagrams. Under these conditions, then, individuals should report a greater use of the

J.Y. Shah, A.W. Kruglanski / Journal of Experimental Social Psychology 38 (2002) 368–383

presented means when solving the anagrams; and whereas in Study 1, the priming manipulation occurred prior to the actual anagram task while the participants were working their way through the anagram samples, in the present study, it took place concomitantly with the anagram task. Method Participants Forty students of the University of Wisconsin-Madison participated in the study in exchange for extra credit toward a course requirement. Of these, 6 were male and 34 female. Participants’ sex had no significant effects on any of the present dependent variables and will not be discussed further. Procedure Participants completed a majority of the procedure on IBM-compatible computers. As in Study 1, they were told they would be completing two separate tasks, the first of which would involve the solution of anagrams. Participants were told that following the anagram task that assesses ‘‘verbal recall’’ they would be completing the box-use task assessing ‘‘functional thinking.’’ Participants were also presented with an alleged strategy for solving anagrams. It was referred to as the ‘‘pair-letter’’ strategy and it involved searching for common letter pairings in each anagram (e.g., the ‘‘ch’’ in ‘‘cpunh’’). Pre-testing had indicated that use of this strategy actually had no effect on the anagram performance for the presented set of anagrams. As in Study 1, participants were also told that individuals with good verbal recall could find 80% of the possible words in the set of anagrams and were first given some sample anagrams before completing the actual anagram task. After completing the sample anagrams, participants rated how difficult and enjoyable they found the anagram task, the importance of the box-use task, and the perceived effectiveness of the ‘‘pair-letter’’ strategy. Participants were also asked to rate how related they thought verbal recall was to functional thinking. Although these two tasks again differed in terms of their contents, it was assumed that their close placement together in the same study might cause some participants to infer that these tasks were related to each other. We had therefore again assumed a natural variability in our sample on the tasks’ perceived relatedness. This expectation was confirmed. On a 10-point scale ranging from 1 (not at all) to 10 (extremely), participants’ relatedness ratings varied from 3 to 10, with a median score of 6. Participants then completed a set of 8 anagrams, comprising the anagram task. Before presenting each anagram trial, we subliminally primed the participants with a word phrase. In the two low-accessibility condi-

373

tions, participants were presented with the control phrase ‘‘view it’’ for 50 ms backward masked in the same manner described previously. In the two high-accessibility conditions, participants were presented with the phrase ‘‘box-use’’ for 50 ms backward masked in the same manner as in Study 1. No participants reported seeing the presented word phrases. The computer recorded the answers that participants provided for each anagram. After completing the entire anagram series, participants were asked to indicate how frequently they had used the ‘‘pair-letter’’ strategy while completing the anagram task and how often they had thought about the box-use task while completing the anagram task. The latter question was included as a check on whether our priming manipulation affected accessibility of the alternative goal in the manner that we had intended. Participants’ responses were recorded on the same 10-point scale described earlier. Finally, a funneled debriefing was used to assess whether participants had guessed the true nature of the study. This debriefing was identical in format to the one described in Study 1. Again, no participants reported any suspicions about the study before the debriefing nor did any report seeing the word primes after being told that they occurred during the task. Finally, participants were thoroughly debriefed about the true nature of the study and invited to ask any remaining questions about its purpose. Results Relation among variables An initial correlational analysis found that participants’ rating of the facilitative relation between the two tasks was not related to how enjoyable they found the task, r ¼ :09, p > :50, but was marginally related to how difficult they found it, r ¼ :29, p < :10, and to the effectiveness of the presented attainment means, r ¼ :28, p < :10 as well as significantly related to how committed participants were to the box-use task, r ¼ :32, p < :05. This correlational analysis also found participants’ persistence on the anagrams to be significantly positively correlated with their performance, r ¼ :49, p < :001. As in Study 1, participants’ ratings of the facilitative interrelation of the two task goals they were expecting to pursue were subjected to a median split to form separate high relation and low relation groups. As in Study 1, we also examined whether the accessibility manipulation affected the perceived difficulty of the anagram task and the importance of the box-use task and found that it had no significant effect on either rating, F s < 1. Check on the accessibility manipulation We then proceeded to examine whether our accessibility manipulation affected the degree to which partic-

J.Y. Shah, A.W. Kruglanski / Journal of Experimental Social Psychology 38 (2002) 368–383

r

374

Fig. 2. Anagram performance and persistence as a function of alternative priming and perceived relation (Study 2).

ipants reported thinking about the box-use task while completing the anagram task. Because the accessibility manipulation also affected anagram persistence and because the time participants spent thinking about the box-use task would be expected to relate to the total time they spent on the anagram task, we controlled in this ANOVA for overall differences in time spent on the anagram task as well as participants’ other task ratings. As predicted, participants reported that they thought about the box-use task significantly more in the high accessibility condition than in the low-accessibility condition, F ð1; 31Þ ¼ 4:85, p < :05. The perceived interrelation of the two tasks exerted no main or interactive effects on this rating, F s < 1. Thus, it appears that our subliminal priming manipulation was successful. Even though it was consciously undetected by our participants, it significantly impacted the degree to which they reported thinking about the primed alternative task goal. Anagram performance and persistence In separate ANOVAs, we then examined the effect of accessibility manipulation on participants’ anagram performance and persistence for individuals who perceived the goals to be related and for those who perceived them to be unrelated, controlling for their other task ratings. Participants in the high-accessibility condition did not perform significantly worse overall (although the effect approached significance, F ð1; 32Þ ¼ 2:43, p > :10), but they did quit marginally sooner than participants in the low-accessibility condition, F ð1; 32Þ ¼ 2:78, p ¼ :10. These analyses also revealed that perceived goal interrelation, as such, did not have a significant effect on the anagram performance or persistence, F ð1; 32Þ ¼ 1:72, p > :15; F < 1, but it did significantly interact with accessibility, (F ð1; 32Þ ¼ 5:61,

p < :05; F ð1; 32Þ ¼ 5:46, p < :05, respectively), indicating that the effect of the alternative goal’s accessibility on anagram performance and persistence was significantly greater when the perceived interrelation between the goals was low than when it was high. Participants’ total anagram performance and untransformed persistence (in seconds) as a function of type of prime and perceived inter-goal relation are presented in Fig. 2. These results replicate the findings of Study 1, attesting to their robustness. Provided means use In a final ANOVA, we analyzed the effect of an alternative goal’s accessibility and its perceived relation to the focal goal on the degree to which participants reported using the means we had provided them for working on the anagrams. Again, there was a significant interaction between the accessibility and interrelation variables, suggesting that participants were significantly more likely to rely on the presented means (rather than generating independent means) when the alternative goal was accessible and unrelated to the focal goal, F ð1; 32Þ ¼ 9:56, p < :01. This interaction is illustrated in Fig. 3. As can be seen, the self-reported use of the provided means is by far greater when the primed term represents an unrelated alternative versus a control word. However, the self-reported use of the provided means is not appreciably different when the prime represents a facilitatively related goal versus a control word. In summary, the results of Study 2 replicate those of Study 1, demonstrating the robustness of our initial findings across several modifications in procedure and extend them by considering how accessibility effects may influence the use of attainment means. Once again the

J.Y. Shah, A.W. Kruglanski / Journal of Experimental Social Psychology 38 (2002) 368–383

375

Fig. 3. Use of provided means as a function of alternative priming and perceived relation (Study 2).

accessibility of alternative goals was found to negatively impact goal pursuit (as measured by anagram performance and persistence), this effect being particularly pronounced when the alternative goal was perceived as facilitatively unrelated (vs. related) to the anagram task. Study 2 also validated our accessibility manipulation by demonstrating that it affected the degree to which participants thought about the alternative goal (the box-use task) while completing the anagrams. Finally, and most importantly, Study 2 confirmed our prediction about the effects of accessible alternative goals on reported reliance on externally provided means of goal attainment. As hypothesized, reported reliance was greatest when the primed alternative goal was facilitatively unrelated (vs. related) to the focal goal. But can one be sure that this self-reported reliance reflected actual means use? Study 3 was specifically designed to investigate this issue.

Study 3 Method Participants Thirty-five students of the University of WisconsinMadison participated in the study in exchange for extra credit toward a course requirement. Of these, 17 were male and 18 female. Participants’ sex had no significant effects on any of the present dependent variables and will not be discussed further. Procedure Participants completed a majority of the procedure on IBM-compatible computers. The general experimental paradigm was similar in most respects to that of Studies 1 and 2. Participants were informed that they would be completing two separate tasks. The first of

these tasks would involve the solving of anagrams and was characterized as a measure of ‘‘verbal fluency.’’ Participants were again told that after completing the anagram task they would be completing the box-use task assessing their ‘‘functional thinking.’’ As in Study 2, participants in the present study were also presented with a strategy for solving the anagrams. However, this strategy was different from that presented in Study 2. Whereas in Study 2, participants were provided with the ‘‘pair-letter’’ strategy, and in Study 3, participants were provided the ‘‘first–last’’ strategy, which involved determining whether any of the possible solutions started with the first or last letter of the anagram. Each of the anagrams had at least one solution that could be found through the use of this particular strategy and one solution that could not be found through its use. This design feature was implemented to allow us to determine the actual use of the supplied strategy, as evident in the ‘‘mean-related’’ solutions that participants found and their use of other strategies, as evident in the solutions they found that required the use of different means than the one provided. All participants were then told that individuals with good verbal fluency could find 80% of the possible words in the set of anagrams and were first given a sample of anagrams before attempting the actual anagram task. After completing the sample anagrams, participants rated how difficult and enjoyable they found the anagram task, the importance of the box-use task, the perceived effectiveness of the ‘‘first–last’’ strategy, and the perceived facililitative relation between functional thinking and verbal fluency. Each rating was performed on a 10-point scale ranging from 1 (not at all) to 10 (extremely). Again, participants’ perceptions of the tasks’ facilitative relatedness exhibited a natural variability across the entire range of the scale. Specifically, they varied between the scores of 1 and 10, with a median score of 6. Participants then performed a set of 8

376

J.Y. Shah, A.W. Kruglanski / Journal of Experimental Social Psychology 38 (2002) 368–383

anagrams, comprising the anagram task. Before each anagram was presented, participants were subliminally primed with a word phrase. In the low-accessibility condition, they were presented with the control phrase ‘‘see this’’ for 50 ms backward masked using the same mask as described in Study 1 (‘‘xxxxxxxxxxxxxx’’). In the high-accessibility condition, they were presented with the phrase ‘‘box use’’ for 50 ms backward masked in the same manner. No participants reported noticing the presented word phrases. The computer recorded the answers participants provided for each anagram. After participants completed the entire anagram series, a funneled debriefing identical to that of Study 1 was implemented to assess whether participants had guessed the true nature of the present experiment. No participants reported any suspicions about the study before the debriefing nor did any report having seen the word primes after being told that they had been presented on the screen. Results

Anagram performance and persistence As in Studies 1 and 2, we then performed separate ANOVAs to examine the effect of the accessibility manipulation and perceived inter-goal relation on participants’ anagram performance and persistence for individuals who perceived the goals to be facilitatively related and those who perceived them to be unrelated, controlling for participants’ remaining ratings of the task. Neither the accessibility manipulation nor the measure of perceived goal interrelation exerted main effects on anagram performance (F s < 1) and persistence (F < 1, and F ð1; 27Þ ¼ 1:90, p > :15, respectively). However, they did so interactively such that the detrimental effects of an alternative goal’s accessibility on performance and persistence were significantly reversed when perceived goal interrelation was high versus being low, F ð1; 27Þ ¼ 5:34, p < :05; F ð1; 27Þ ¼ 4:62, p < :05, respectively. Participants’ total anagram performance and untransformed persistence time (in seconds) as a function of type of prime and perceived inter-goal relation are presented in Fig. 4. These results replicate the findings of the first two studies. Actual means use Of greater interest in the present experiment were the possible effects of our variables on the actual use of the provided means for anagram solution. Participants could have found a number of the possible solutions to the anagrams by using the provided

r

Relation among variables An initial correlational analysis found that participants’ rating of the facilitative relation between the two tasks was not related to how difficult or enjoyable they found the anagram task or how important they viewed the box-use task (r ¼ :12, p > :45, r ¼ :03, p > :80; r ¼ :12, p > :50, respectively) but significantly related to the perceived effectiveness of the presented attainment means, r ¼ :37, p > :05. This correlational analysis also found participants’ persistence on the anagrams to be significantly positively correlated with their performance, r ¼ :45, p < :005. As in our prior studies, participants’ ratings of the interrelation of the task goals were subjected to a

median split to form separate high relation and low relation groups. Also consistent with our previous two experiments, the present accessibility manipulation did not appear to affect the perceived difficulty of the anagram task or the importance of the box-use task, F s < 1.

Fig. 4. Anagram performance and persistence as a function of alternative priming and perceived relation (Study 3).

J.Y. Shah, A.W. Kruglanski / Journal of Experimental Social Psychology 38 (2002) 368–383

377

Fig. 5. Anagram performance as a function of alternative priming, perceived relation, and whether attainment means was provided (Study 3).

means ( the ‘‘first–last’’ strategy). A comparable number of possible solutions to these anagrams, however, required participants to use different means than the one provided. To determine the degree to which participants relied on the given means and the degree to which they generated their own means, we calculated the total number of solutions found by participants that involved using the provided means and the total number of solutions that they found using different means. Each total was examined as a function of the accessibility of the alternative and perceived goal interrelations, as summarized in Fig. 5. As can be seen, when the alternative goal was accessible (through priming) and seen as facilitatively unrelated to the focal goal, participants found more of the solutions that involved using the provided means than the solutions that involved using different means. When the alternative was accessible and seen as facilitatively related, however, participants found more solutions that involved using different means than solutions involving the use of the provided means. Performing a MANOVA, we found that the overall interaction of the accessibility manipulation, perceived goal interrelation, and type of solutions found (provided means vs. different means) was significant (F ð1; 27Þ ¼ 5:20, p < :05). This suggests that when the alternative was made more accessible through priming and was seen as unrelated to the focal goal, participants did not generate effective anagram solution strategies on their own and instead tended to rely on the provided means, even though not all solutions to the anagrams could be found by using this strategy. In contrast, when the alternative goal was

made more accessible and seen as facilitatively related to the focal goal, participants found more solutions from the use of different means than from the use of the provided means, presumably because they more readily generated their own strategies for finding solutions to the anagrams. Thus, the pulling of resources away from the focal goal—when the accessible alternative goal was facilitatively unrelated to it—may have lowered participants’ quest for effective means of goal attainment, which, in turn, may have contributed to the observed deterioration of anagram performance in the accessibleunrelated condition. Likewise, the mobilization of resources in service of the focal goal when the alternative was facilitatively related thereto may have increased participants’ quest for effective means for goal attainment, contributing, in turn, to the observed increase in anagram performance in the accessible-related condition. Our final study sought to replicate the general pattern of performance and persistence found in the first three studies using a different alternative goal and to extend the findings of these studies by considering participants’ emotional reactions to feedback about progress or the lack of progress on the focal task. Our theoretical analysis suggests that commitment to the focal goal should be affected by the accessibility of an alternative goal and its relation to the focal pursuit. These variables should therefore exert a parallel influence on participants’ positive or negative affective reactions to success or failure-feedback regarding this focal pursuit. Participants whose goal commitment is undermined by the presence of alternative goals should

378

J.Y. Shah, A.W. Kruglanski / Journal of Experimental Social Psychology 38 (2002) 368–383

care less and hence be less emotionally reactive to such feedback.

Study 4 Method Participants Ninety students of the University of WisconsinMadison participated in the study in exchange for extra credit toward a course requirement. Of these, 38 were male and 52 female. Participants’ sex did not significantly affect any of the present dependent variables and will not be discussed further. Procedure As in our previous studies, participants completed a majority of the procedure on IBM-compatible computers. They were told that they would be completing two separate tasks, the first of which would involve the solution of anagrams. This initial task was described as a measure of ‘‘verbal reasoning,’’ a valuable skill for writing. Unlike the alleged ‘‘functional thinking’’ task used in our prior studies, a different second task was introduced and described as a measure of ‘‘analytic reasoning.’’ We again assumed that some participants, at least, would view the two tasks as facilitatively related, as they both pertained to ‘‘reasoning’’ and both required participants to generate as many different solutions as possible. Participants were told that this second task would require them to combine a set of four numbers to form a fifth number by adding, subtracting, multiplying, and dividing the set of numbers as well as using parentheses. For instance, ‘‘ð4  3Þ þ 2  1’’ would be one way to get the number set ‘‘4 3 2 1’’ to equal ‘‘13.’’ After the two tasks were described to participants, they completed three practice anagrams and three practice trials of the ‘‘analytic reasoning’’ task. They then proceeded with the anagram task. As with all our previous studies, this task was the only one described that participants actually attempted. Prior to embarking on the anagram task, participants were informed that individuals with good verbal reasoning skills could find 80% of the possible words in the set of anagrams. They were then asked, ‘‘how difficult does the anagram task seem?’’ and ‘‘how enjoyable does the anagram task seem?’’ and were also asked to rate how satisfied, calm, dejected, and anxious they were feeling at that moment. Answers to each of these questions were provided on a 7-point scale ranging from 1 (not at all) to 7 (extremely). Participants then completed the anagram task, which consisted of 10 trials in this study. The computer recorded the answers participants provided for each anagram and the amount of time they spent on each trial.

Before each anagram’s appearance in the low-accessibility condition, participants were presented with the word ‘‘dedication’’ for 50 ms backward masked in the same manner described in Study 1. In the high-accessibility conditions, they were presented with the word ‘‘analytic’’ in the same fashion. No participant in either condition reported having seen the presented word. Participants were randomly assigned to one of the two feedback conditions. After completing the anagrams, those in the success-feedback condition were informed that they had, in fact, found 80% of all the possible solutions. Participants in the failure-feedback condition were told that they had not found 80% of all the possible solutions. After receiving the feedback, participants again rated how satisfied, calm, dejected, and anxious they were feeling at that moment. Finally, participants were asked to rate how related verbal reasoning was to analytic reasoning and how important it was for them to do well on the upcoming analytic reasoning task. Each of these ratings was given on the same 7-point scale described earlier. As expected, participants’ actual facilitation ratings naturally varied over a considerable range, namely from 1 to 7, with a median score of 4. A funneled debriefing identical in format to that of Study 1 was then used to assess whether participants had guessed the true nature of the experiment. Again, no participants reported any suspicions before the debriefing nor did any report noticing the word primes after having been told that they were flashed on the screen. Affective reactivity After reverse scoring all the negative mood items, participants’ change in mood was calculated by summing their mood scores after receiving feedback and subtracting from it the sum of their initial mood scores. This score represented the magnitude of their emotional reaction to success or failure-feedback. Results Relation among variables An initial correlational analysis found that participants’ rating of the facilitative relation between the two tasks was not related to how enjoyable they found the anagram task, r ¼ :08, p > :45, and how important they found the analytic reasoning task, r ¼ :08, p > :45, but was marginally negatively related to how difficult they found the anagram task, r ¼ :19, p < :10. This correlational analysis also found participants’ persistence on the anagrams to be significantly positively correlated with their performance, r ¼ :40, p < :001. Because, in this study, alternative accessibility and feedback were manipulated before participants rated the facilitative relation between the two tasks, we conducted a separate ANOVA to examine the possi-

J.Y. Shah, A.W. Kruglanski / Journal of Experimental Social Psychology 38 (2002) 368–383

bility that these manipulations affected participants’ facilitative relation rating. No evidence was found for this possibility, as the ANOVA revealed that participants’ facilitative relation rating was not significantly affected by the alternative accessibility manipulation, the feedback manipulation, or the interaction of these manipulations (all F s < 1). As in the previous studies, participants’ rating of the two tasks’ interrelation was subjected to a median split to form separate high interrelation and low interrelation groups. Conceptually replicating our previous results, separate ANOVAs found that the present accessibility manipulation did not affect the perceived difficulty of the anagram task or the importance of the analytic reasoning task, F s < 1. Effects on anagram performance and persistence Replicating the results of our previous studies and controlling for participants’ task difficulty, task enjoyment, and importance ratings, separate ANOVAs found that the accessibility manipulation and the perceived interrelation of the goals did not significantly affect anagram performance and persistence (all F s < 1) but they did so in interaction with each other (F ð1; 83Þ ¼ 4:33, p < :05; F ð1; 83Þ ¼ 4:32, p < :05, for performance and persistence, respectively). Once again, the detrimental effect of alternative goal accessibility on anagram performance and persistence was significantly greater when perceived task interrelation was low than when it was high. Participants’ total anagram performance and untransformed persistence time (in seconds) as a function of prime-type and perceived inter-goal relation are presented in Fig. 6. These results replicate the findings of our previous studies in this series. Affective reactivity A separate ANOVA examined participants’ emotional response to feedback as a function of alternative goal accessibility and perceived inter-goal relation,

379

controlling for participants’ task ratings. The analysis revealed that feedback had a strongly significant effect on change in mood, F ð1; 79Þ ¼ 42:95, p < :001. Not very surprisingly, positive feedback led to a more positive change in mood than did negative feedback. Neither the accessibility of the alternative nor the perceived interrelation of the goals exerted significant main effects on mood change, F s < 1. Nor did these two variables interact with each other (F < 1) or with feedback to affect the change in mood, F < 1; F ð1; 79Þ ¼ 1:48, p > :20, respectively. However, the three-way interaction of the accessibility of the alternative, the perceived interrelation of the goals and feedback was significant, F ð1; 79Þ ¼ 5:63, p < :05, indicating that the effect of feedback was moderated by the interaction of alternative accessibility and perceived interrelation of the goals. These results are portrayed in Fig. 7. As can be seen, when the alternative goal is facilitatively unrelated to the focal goal, its priming dampens, by and large, participants’ affective reactivity to feedback with regard to progress toward the focal goal. In contrast, when the alternative goal is perceived as related to the focal goal, its priming enhances the affective reactivity to feedback about progress.

General discussion The present research In his analysis of goal conflict, Lewin (1935, 1951) likened goals to physical bodies occupying specific positions within an individual’s ‘‘life space.’’ Each entity in this field either attracts or repels one’s focus and the strength of these attractive and repellent forces varies as a function of the goal’s psychological distance from the individual’s self. The present research emphasizes how the force exerted on an individual by any one goal

Fig. 6. Anagram performance and persistence as a function of alternative priming and perceived relation (Study 4).

380

J.Y. Shah, A.W. Kruglanski / Journal of Experimental Social Psychology 38 (2002) 368–383

Fig. 7. Change in mood as a function of alternative priming, perceived relation, and feedback (Study 4).

cannot be considered in isolation and instead depends on the salience and interrelation of other goals present in one’s life space at a given moment. Four studies consistently suggest that priming alternative goals can affect one’s commitment to a currently pursued objective, which, in turn, has implications for goal-directed behavior and emotion. Moreover, such priming effects are consistently moderated by the degree to which the goal facilitatively relates to its alternative. These conclusions are supported by a convergence of findings pertaining to different accompanying aspects of goal pursuit. Given the relative novelty of our approach to goals and their interrelations, it is encouraging to have replicated, time and time again, the manner by which commitment and progress toward a focal goal is affected by the accessibility of an alternative goal. We find that the extent to which an alternative goal is rendered accessible affects participants’ performance and persistence in striving for the focal goal and we interpret these effects as reflecting the waxing and waning of goal focus that results from the subliminal presentation of a facilitating or unrelated alternative goal. This presentation either pulls attention and resources away from the focal goal (and toward the alternative), or draws attention and resources toward the focal goal by increasing its accessibility. Consistent with such an interpretation, we find that emotional responses to goal pursuit are amplified or dulled by the salience of an alternative goal and its facilitative relation to the focal pursuit. Presumably, these effects reflect the pulling of resources away or the drawing of them toward the consciously pursued goal. We find that the development of efficient means to the focal goal, an obvious resource-intensive activity, is also af-

fected when alternative goals are activated. It is particularly noteworthy that our effects obtain, even though our priming manipulations were subliminal and our participants reported no conscious awareness of their occurrence. This attests to the relatively automatized, non-deliberative nature of the ‘‘goal pull’’ effects. Alternative explanations of goal pull Our results also provide evidence that counters two major alternative explanations to how a focal pursuit is undermined by an unrelated and cognitively salient alternative goal. One such explanation holds that the undermining occurs due to a mere distraction rather than due to the pull on resources by the alternative unrelated goal. Yet contrary to this general distraction hypothesis, we did not find similar effects when priming control constructs. Thus, our findings could not be due to the priming of cognitive constructs per se, but rather due to the specific priming of goal constructs. Moreover, such a hypothesis could not explain the kinds of moderation of the goal pull effects that depend on how an alternative relates to the present pursuit. A second explanation holds that the priming manipulation enhances the perceived importance of the activated alternative goal and/or the perceived difficulty of the focal goal. It is such shift in perception rather than a change in commitment to the focal goal that might have caused our various effects. Again, however, our results seem to belie these possibilities in that neither the perceived importance of the alternative goal nor the perceived difficulty of the focal goal was significantly affected by our priming manipulations in any of our studies.

J.Y. Shah, A.W. Kruglanski / Journal of Experimental Social Psychology 38 (2002) 368–383

Self-regulatory implications The present cognitive and motivational analysis of goal effects has implications for a wide range of significant intra- and inter-personal phenomena. Two such implications are particularly noteworthy. They concern (a) the resolution of goal conflict and (b) the quality of one’s social relations. Implications for the resolution of goal conflict An important question to pose at this point is how one might cope with the conflict occasioned by the activation of alternative goals rival to one’s current pursuits. When Buridan was asked to explain how an ass, or anyone else, might overcome the decisional challenge posed by two equally attractive alternatives, he had a ready answer: one could consider a particular alternative closely and thereby cause any others to fall out of conscious awareness. Such a solution, with its emphasis on lowering the accessibliity of alternatives, can readily be applied to the resolution of goal conflict. It suggests that one’s ability to cognitively inhibit, or actively put aside, alternative goals may allow one to focus more resources on a focal pursuit. It is precisely such processes that may underlie the observed individual differences in coping with competing or conflicting goals in social and academic settings. Locke and Kristof (1996), for instance, found that individuals differ in the degree to which they are able to ignore the attractiveness of other pursuits and that this difference positively predicts their college grade point average. Similarly, Rabiner and Gordon (1992) found that aggressive and rejected children were less able to manage the simultaneous and often conflicting goals inherent in many social situations. Dodge and Crick (1984) have suggested that children’s social competence is best understood not by the specific goals they tend to adopt in social situations, but by their ability to coordinate their goals with the goals of others and to manage the inevitable conflicts that arise. They suggest that prioritizing or deferring goals are important tools in this coordination but when attempting it, socially incompetent children are often unable to ignore the attractiveness of their own goals or see their relation to the goals of others. The present results offer insights into possible underlying mechanisms behind these individual differences. Of particular interest in this regard is the relatively automatic nature of the ‘‘goal pull’’ effects, suggesting that to counter them one may need to automatize the inhibition of alternative goals as well. If automatization develops in the context of practice and experience (cf. Bargh, 1996) it may be necessary to institute just such an inhibitive experience to improve persons’ ability to

381

focus on their tasks. These possibilities appear to merit the investment of further research resources in their exploration. Implications for inter-personal relations A final, though by no means least important, implication of our analysis has to do with the quality of our relationships to significant others. Mention has been made already of the fact that other persons could often trigger various alternative goals to one’s current pursuits. In addition, some persons could themselves embody a variety of alternative goals, hence, constituting ‘‘hubs’’ wherein significant goals may meet. For instance, John’s wife Ann could represent to him at once the goals of ‘‘romantic love,’’ ‘‘giving and receiving affection,’’ ‘‘camaraderie and social support,’’ ‘‘securing the children’s education,’’ ‘‘paying the bills,’’ ‘‘cleaning the house,’’ etc. Obviously, different social targets may vary in the number of goals with which they are associated and in the strength of those associations as well. This, in turn, may determine how one comes to feel about and interact with the person in question. Researchers on close relationships have commented on the decline of romance and passion in marital relations as time goes by (i.e., Berscheid, Synder, & Omoto, 1989; Huston & Chorost, 1994). From the present perspective, one reason for this decline could be a dilution of the romantic/passionate commitment as the marital partners come to represent to each other a bundle of simultaneously activatable and unrelated goals competing for their mental and physical resources. In short, the accessibility of alternative goals may carry profound psychological implications not only for motivational and cognitive (e.g., attributional) phenomena at the intra-individual level but also for inter-personal relations and social interaction.

Conclusion Although traditionally, motivational and cognitive explanations of psychological phenomena have tended to run along separate paths (see Bem, 1972; Miller & Ross, 1975), there has been a growing appreciation of the cognitive aspects of motivational constructs such as goals (cf. Bargh & Barndollar, 1996; Kruglanski, 1996a,b). The present set of studies strengthens this merger by illustrating how a cognitive quality of alternative goals (their accessibility) and a motivational quality (their perceived relation to a focal pursuit) interact to affect the singularity and strength with which one strives toward a focal goal, as well as one’s emotional reaction to its attainment or attainment failure. By considering the cognitive qualities of goals and how

382

J.Y. Shah, A.W. Kruglanski / Journal of Experimental Social Psychology 38 (2002) 368–383

they interact with goals’ strictly motivational aspects, the present research has begun to demonstrate the considerable potential of a ‘‘warm look’’ on motivation (see Sorrentino & Higgins, 1986). Such a perspective not only recognizes the distinct motivational and cognitive facets of goal pursuit, but also highlights their intricate inter-connections.

References Bargh, J. A. (1990). Auto-motives: Preconscious determinants of social interaction. In E. T. Higgins, & R. M. Sorrentino (Eds.), Handbook of motivation and cognition (Vol. 2) (pp. 93–130). New York: Guilford Press. Bargh, J. A. (1996). Automaticity in social psychology. In E. T. Higgins, & A. W. Kruglanski (Eds.), Social psychology: Handbook of basic principles (pp. 169–183). New York: Guilford. Bargh, J. A., & Barndollar, K. (1996). Automaticity in action: The unconscious as repository of chronic goals and motives. In P. M. Gollwitzer, & J. A. Bargh (Eds.), The psychology of action (pp. 457– 481). New York: Guilford. Bargh, J. A., Chaiken, S., Govender, R., & Pratto, F. (1992). The generality of the automatic attitude activation effect. Journal of Personality and Social Psychology, 62, 893–912. Bargh, J. A., & Gollwitzer, P. M. (1994). Environmental control of goal-directed action: Automatic and strategic contingencies between situations and behavior. In W. D. Spaulding (Ed.), Integrative views of motivation, cognition, and emotion (pp. 71–124). Lincoln: University of Nebraska Press. Bem, D. J. (1972). Self-perception theory. In L. Berkowitz (Ed.), Advances in experimental social psychology (Vol. 6) (pp. 1–62). New York: Academic Press. Bem, D. J. (1967). Self-perception: An alternative interpretation of cognitive dissonance phenomena. Psychological Review, 74, 183– 200. Berscheid, E., Synder, M., & Omoto, A. M. (1989). Issues in studying close relationships: Conceptualizing and measuring closeness. In C. Hendrick (Ed.), Review of Personality and Social Psychology (Vol. 10). Newbury Park, CA: Sage. Brunstein, J. C., & Olbrich, E. (1985). Personal helplessness and action control: An analysis of achievement-related cognitions self assessments, and performance. Journal of Personality and Social Psychology, 48, 1540–1551. Chartrand, T. L., & Bargh, J. A. (1996). Automatic activation of impression formation and memorization goals: Nonconscious goal priming reproduces effects of explicit task instructions. Journal of Personality and Social Psychology, 71, 464–478. Deci, E. L. (1971). Effects of externally mediated rewards on intrinsic motivation. Journal of Personality and Social Psychology, 18, 105– 115. Dodge, K. A., & Crick, N. R. (1984). Social life as a goal coordination task. In C. Ames, & R. Ames (Eds.), Research on motivation in education Goals and cognitions (Vol. 3) (pp. 107–135). San Diego: Academic Press. Higgins, E. T. (1989). Self-discrepancy theory: What patterns of selfbeliefs cause people to suffer? In L. Berkowitz (Ed.), Advances in experimental social psychology (Vol. 22) (pp. 93–136). New York: Academic Press. Higgins, E. T. (1997). Beyond pleasure and pain. American Psychologist, 52, 1280–1300. Higgins, E. T., Shah, J. Y., & Friedman, R. (1997). Emotional responses to goal attainment: Strength of regulatory focus as moderator. Journal of Personality and Social Psychology, 72, 515– 525.

Huston, T. L., & Chorost, A. F. (1994). Behavioral buffers on the effect of negativity on marital satisfaction: A longitudinal study. Personal Relationships, 1, 223–239. Judd, C. M., & McClelland, G. H. (1989). Data analysis: A modelcomparison approach. Orlando, FL: Harcourt Brace Jovanovich. Kruglanski, A. W. (1975). The endogenous–exogenous partition in attribution theory. Psychological Review, 82, 387–406. Kruglanski, A. W. (1996a). Goals as knowledge structures. In P. M. Gollwitzer, & J. A. Bargh (Eds.), The psychology of action: Linking cognition and motivation to behavior (pp. 599–619). New York: Guilford. Kruglanski, A. W. (1996b). Motivated social cognition: Principles of the interface. In E. T. Higgins, & A. W. Kruglanski (Eds.), Social psychology: A handbook of basic principles (pp. 493–522). New York: Guilford. Kruglanski, A. W., Friedman, I., & Zeevi, G. (1971). The effects of extrinsic incentive on some qualitative aspects of task performance. Journal of Personality and Social Psychology, 39, 606–617. Kuhl, J. (1984). Volitional aspects of achievement motivation and learned helplessness. Toward a comprehensive theory of action control. In B. A. Maher, & W. A. Maher (Eds.), Progress in experimental personality research (pp. 99–171). New York: Academic Press. Kuhl, J., & Beckman, J. (Eds.). (1994). Volition and Personality. Toronto: Hogrefe. Kuhl, J., & Helle, L. (1986). Motivational and volitional determinants of depression: The degenerated-intention hypothesis. Journal of Abnormal Psychology, 95, 247–251. Kuhl, J., & Kazen-Saad, M. (1988). A motivational approach to volition: Activation and de-activation of memory representations related to uncompleted intentions. In V. Hamilton, G. H. Bower, & N. H. Frijda (Eds.), Cognitive perspectives on emotion and motivation (pp. 63–85). Dodrecht: Kluwer Academic Publishers. Kuhl, J., & Weiss, M. (1985). Performance deficits following uncontrollable failure: Impaired action control or generalized expectancy deficits? Munich: Max Planck Institute for Psychological Research, Paper 5/84. Kunda, Z. (1990). The case for motivated reasoning. Psychological Bulletin, 108, 480–498. Lepper, M. R., Greene, D., & Nisbett, R. E. (1973). Undermining children’s intrinsic interest with extrinsic reward: A test of the overjustification hypothesis. Journal of Personality and Social Psychology, 28, 129–137. Lewin, K. (1935). A dynamic theory of personality: Selected papers (D.E. Adams K.E. Zener, Trans.). New York: McGraw Hill. Lewin, K. (1951). Field theory in social science. New York: Harper. Locke, E. A., & Kristof, A. L. (1996). Volitional choices in the goal achievement process. In P. M. Gollwitzer, & J. A. Bargh (Eds.), The psychology of action: Linking cognition and motivation to behavior (pp. 365–384). New York: Guilford Press. Miller, N. E. (1944). Experimental studies in conflict. In J. M. Hunt (Ed.), Personality and the behavioral disorders (Vol. 1) (pp. 431– 465). New York: Roland Press. Miller, D. T., & Ross, M. (1975). Self-serving biases in the attribution of causality: Fact or fiction? Psychology Bulletin, 82, 213–225. Mischel, W., & Ebbeson, E. B. (1970). Attention in delay of gratification. Journal of Personality and Social Psychology, 16, 239–337. Mischel, W., Ebbeson, E. B., & Zeiss, A. R. (1972). Cognitive and attentional mechanisms in delay of gratification. Journal of Personality and Social Psychology, 21, 204–218. Mischel, W., Shoda, Y., & Rodriguez, M. L. (1989). Delay of gratification in children. Science, 244, 933–938. Rabiner, D. L., & Gordon, L. V. (1992). The coordination of conflicting social goals: Differences between rejected and nonrejected boys. Child Development, 63, 1344–1350. Shah, J.Y., Kruglanski, A.W. (2000c) When the appetite comes from the eating: Bottom-up priming of goals by means and the

J.Y. Shah, A.W. Kruglanski / Journal of Experimental Social Psychology 38 (2002) 368–383 implications for self-regulation. (Manuscript submitted for publication). Shallice, T. (1972). Dual functions of consciousness. Psychological Review, 79, 383–393. Sorrentino, R. M., & Higgins, E. T. (Eds.). (1986). Handbook of motivation and cognition: Foundations of social behavior. New York: Guilford. Turner, W. (1903). History of philosophy. Boston: Ginn and Company.

383

Trzebinski, J. (1989). The role of goal categories in the representation of social knowledge. In L. Pervin (Ed.), Goal concepts in personality and social psychology (pp. 363–411). Hillsdale: Erlbaum. Zeigarnik, B. (1938). On finished and unfinished tasks. In W. D. Ellis (Ed.), A source book of gestalt psychology (pp. 300–314). New York: Harcourt, Brace and World (Reprinted and condensed from Psychologische Forschung, 9, 1–85, by B. Zeigarnik, 1927).