*Ji Kyung Park is Assistant Professor of Marketing, Alfred Lerner Col- lege of Business and Economics, University of Delaware (e-mail; jipark@ udel. edu). Deborah Roedder John is Curtis L. Carlson Chair and Professor of Marketing, Carlson School of Management, University of Minnesota (e-mail: djohn@umn.edu). Both authors contributed equally to this article. The authors acknowledge many helpful suggestions from the anonymous JMR reviewers. Gita Johar served as associate editor for this article. JI KYUNG PARK and DEBORAH ROEDDER JOHN* When consumers struggle with a difficult task, using a brand can help them perform better. The authors report four studies showing that brand use can enhance feelings of self-efficacy, which can lead to better task performance. Students scored higher on difficult Graduate Records Examination questions when they took the test using a Massachusetts Institute of Technology pen (Study 1) and showed better athletic performance when they drank water from a Gatorade cup during strenuous athletic exercise (Studies 2 and 3). These increases in task performance were mediated by feelings of self-efficacy (Studies 3 and 4). Furthermore, the results show that not everyone experiences the beneficial effect of brand use; it depends on the persons implicit self- theory. Across studies, users adopting entity theories (entity theorists) showed increased self-efficacy and better task performance, whereas users adopting incremental theories (incremental theorists) were unaffected by brand use. Keywords: brands, self-efficacy, implicit self-theories I Think I Can, I Think I Can: Brand Use, Self-Efficacy, and Performance 2014, American Marketing Association ISSN: 0022-2437 (print), 1547-7193 (electronic) 233 Consumers face a variety of challenges in their everyday lives. Getting through a tough exercise routine, finishing a difficult work assignment, and disciplining an obstinate child are just a few examples of activities that present an uphill battle for many people. In each case, consumers must persist in the face of obstacles to complete the task. When they fail, it is often because they lack confidence in their abilities. This lack of confidence undermines their determi- nation to work through the obstacles they face. Unfortu- nately, many of us do not live up to the saying When the going gets tough, the tough get going. We propose that brands can help consumers with chal- lenging situations in their lives. Specifically, we suggest that using brands can increase consumers sense of self-efficacy, which refers to the belief in ones capabilities to perform well in a particular situation (Bandura 1995). Self-efficacy is an important regulatory mechanism that governs the level of challenge people are willing to undertake, and it thus determines actual performance on challenging tasks (Bandura 1982; Wood and Bandura 1989). We posit that consumers can feel more confident about their ability to perform well (self-efficacy) and can actually perform better when they use a brand that promises benefits helpful for performing the task. For example, consumers engaged in a tough exer- cise routine can feel more confident about their abilities and actually perform betterif they exercise while using a brand that promises better athletic performance, such as Gatorade or Under Armour. Furthermore, we show that there are individual differ- ences in experiencing this self-efficacy effect. In particular, we examine the implicit self-theory people hold as an important moderating factor (Dweck 2000). We find that people who believe that their personal qualities and abilities cannot be improved by their own efforts (entity theorists) experience a boost in self-efficacy and performance when using a brand during a difficult task. In contrast, people who believe that they can improve their personal qualities and abilities through their own efforts (incremental theorists) do not experience the self-efficacy boost from using brands and do not exhibit better task performance. We present four experiments to explore these effects. In each, participants use a branded product while performing a difficult task, such as taking a graduate school admissions test (while using a Massachusetts Institute of Technology [MIT] pen) or engaging in strenuous athletic exercise (while drinking tap water from a Gatorade cup). The products, such as a pen or water, do not have functional qualities that could account for better task performance. For example, a pen does not have functional qualities that could increase a test takers intelligence level or ability to answer Graduate Records Examination (GRE) questions. This aspect of our study enables us to test the brands effects apart from func- tional qualities that could be falsely attributed to the product itself. It also distinguishes our research setting from tradi- tional placebo studies, in which people have a basis for developing false beliefs about inert substances such as a sugar pill because pills usually have functional qualities that cure medical ailments. Our research opens a new area of inquiry in understand- ing how consumers benefit from using brands. First, brands can provide a sense of self-efficacy. Consumers often lack the confidence to engage in challenges that could improve their lives, such as exercising strenuously or making healthy, home-cooked meals. Self-efficacy is critical for success in these situations, and we find that brands can pro- vide the boost necessary to help many consumers perform well. We also identify the type of consumers (entity theo- rists) who experience such benefits from using brands. For example, we find that entity theorists score higher on a diffi- cult graduate school admissions test when using an MIT pen to take the test and show better athletic performance when drinking tap water from a Gatorade cup during strenuous athletic exercise. Second, we identify a process whereby brands can influ- ence behavior that differs from that detailed in prior research. For example, previous studies have found that incidental exposure to a brand (e.g., brand logos for Apple, Disney) can activate goals and goal-consistent behavior related to the brands image (e.g., being creative, behaving honestly) (Chartrand et al. 2008; Fitzsimons, Chartrand, and Fitzsimons 2008). These effects take place on a noncon- scious level. Although brands can automatically activate goals, when tasks are difficult and challenging, simply pos- sessing a goal to perform well may not be sufficient. A cog- nitive appraisal of personal efficacy is required both to increase the level of challenge that people are willing to undertake and to perform well on the task (Zimmerman, Bandura, and Martinez-Pons 1992). By focusing our inquiry on difficult tasks, we show that brands can affect behaviors on a conscious level by driving feelings of self-efficacy. This distinction is important on a managerial level because it is difficult to communicate nonconscious benefits that brands deliver to consumers, but firms can incorporate con- sciously experienced brand benefits into brand messaging and brand promotions. Third, our research confirms the importance of implicit self-theories in brand research. Prior work has found that these beliefs influence how consumers form attitudes toward branded products (Yorkston, Nunes, and Matta 2010). Research has also shown that the use of brands can change peoples self-perceptions about their personality traits, but this change only occurs for entity (not incremen- tal) theorists (Park and John 2010). In this article, we show that, for consumers who hold entity theory beliefs, brands can raise their sense of self-efficacy and increase their level of performance when facing challenging tasks. Thus, our findings confirm the sensitivity of entity theorists to brand cues and extend the realm of influence from self-perceptions to self-efficacy and task performance. Self-perceptions (e.g., I am athletic) are broad evaluations of the self, whereas self-efficacy (I am confident I can do well on this exercise) is task specific and is a key determinant of performance on specific tasks (Pajares and Schunk 2001). Our article proceeds as follows. First, we provide an overview of self-efficacy and implicit self-theories and dis- cuss how brands can provide a self-efficacy boost and why implicit self-theories moderate this effect. Next, we present the results of four experiments that demonstrate the self- efficacy effect, with brand usage resulting in heightened self-efficacy and better task performance for entity theorists only. Our findings also support the idea that self-efficacy mediates better task performance and that the process takes place on a conscious level. Finally, we discuss the contribu- tions of our findings for branding research and the implica- tions of our findings for marketing strategies and further research. CONCEPTUAL BACKGROUND In this section, we describe the concept of self-efficacy in more detail. We then turn our attention to implicit self- theories and propose that entity theorists draw on different sources of self-efficacy than do incremental theorists. We suggest that opportunities to learn and develop skills are a potent source of self-efficacy for incremental theorists, whereas entity theorists rely on assurances that they can be successful as a source of self-efficacy. We argue that brands can provide such assurances and thus can be a source of self-efficacy, which contributes to better performance for entity theorists. Self-Efficacy Self-efficacy refers to the belief in ones capabilities to perform well in a particular situation (Bandura 1995). Beyond this definition, Bandura and Wood have highlighted three aspects of self-efficacy that are particularly important (Bandura 1988; Bandura and Wood 1989; Wood and Ban- dura 1989). First, self-efficacy involves a mobilization com- ponent; thus, people who have the same skills may perform differently because of their utilization and combination of these skills. Second, self-efficacy is a dynamic construct; therefore, self-efficacy beliefs can change over time as peo- ple acquire new information and experiences during task performance. Third, self-efficacy beliefs are linked to dis- tinct realms of functioning (I can do well on this test) as opposed to representing a global trait (I am smart). According to social cognitive theory, self-efficacy is an important self-regulatory mechanism that governs ongoing motivation and performance (Bandura 1986). Self-efficacy beliefs increase the level of challenge people are willing to undertake, the amount of effort they expend in meeting those challenges, and the level of perseverance in the face of difficulties in completing the task, resulting in better per- formance (Bandura 1982; Wood and Bandura 1989). For example, research has shown that self-efficacy beliefs play an important role in controlling eating disorders (Schneider and Agras 1985), increasing strenuous physical activity (Bandura and Cervone 1983), and improving academic per- formance (Bandura 1993). 234 JOURNAL OF MARKETING RESEARCH, APRIL 2014 Brand Use, Self-Efficacy, and Performance 235 Self-Efficacy and Implicit Self-Theories People develop lay theories regarding the nature of the social world to interpret, predict, and control their social worlds (Lickel, Hamilton, and Sherman 2001). Included in this category are implicit self-theories, which are lay beliefs regarding the malleability of personal qualities. Research has identified two types of implicit self-theories: incremen- tal and entity theories. People who endorse incremental theory (incremental theorists) view their personal quali- ties as malleable and believe they can improve on these qualities by devoting effort to self-improvement. In con- trast, people who endorse entity theory (entity theorists) view their personal qualities as more fixed and do not believe they can change these qualities by devoting effort to self-improvement (Dweck 2000). We propose that these mindsets influence the way people develop a sense of self-efficacy. Self-efficacy beliefs are based on a persons assessment of whether he or she has the ability to perform well on a specific task, and we suggest that entity and incremental theorists develop confidence about their abilities in different ways. Incremental theorists are more confident about their abilities when they have opportunities to learn and develop their skills. Because incremental theorists believe they can improve their abilities if they try, they welcome opportunities to learn and develop their skills. Moreover, when they engage in these activities, they gain confidence in their abilities (self-efficacy), which increases performance for tasks requiring these abilities (Dweck 2000; Martocchio 1994; Robins and Pals 2002; Wood and Bandura 1989). However, this is not the case for entity theorists. Entity theorists believe that their abilities are fixed and cannot be improved on through their own efforts, so they show help- less and defensive reactions when given an opportunity to learn and practice new skills. As a result, these opportunities do not increase self-efficacy for entity theorists (Martocchio 1994; Rhodewalt 1994; Robins and Pals 2002). Instead, we propose that entity theorists are more confi- dent about their abilities when they receive assurances that they can perform well on the task. Because entity theorists believe that their abilities are fixed and cannot be improved even if they try, they prefer situations in which they are assured of doing well (Dweck and Leggett 1988; Elliott and Dweck 1988). For example, college students who are entity theorists prefer easier courses in which there is an assurance of getting an A rather than more challenging courses in which they could learn more but might receive a C. Incre- mental theorists exhibit the opposite behavior, choosing more challenging courses even though they may receive a lower grade (Dweck and Leggett 1988; Elliott and Dweck 1988). Thus, entity theorists feel a lack of confidence about their abilities (low self-efficacy) when they face a challeng- ing taskand, unless they receive assurances that they can perform well, entity theorists tend to do poorly in the task (Robins and Pals 2002). Brands as a Source of Assurance Brands can provide assurances of success to consumers through brand promises. A brand promise is a statement of the benefits a brand offers to consumers. Many of these benefits relate to better performance on a task, such as get- ting rid of stubborn stains (Tide) or learning a foreign lan- guage (Rosetta Stone). Gatorade, for example, promises more endurance and better athletic performance, which assures consumers that they can perform better when engaged in a challenging workout routine or sport if they use Gatorade. We predict that entity theorists will use brand promises as a source of self-efficacy. When engaged in a challenging task and using a brand that promises better performance on the task, entity theorists will rely on this promise to increase their confidence about performing well on the task. This increase in confidence will result in better task performance. For example, entity theorists using the Gatorade brand dur- ing a challenging workout routine can find assurance in the brands promise of better athletic performance, which increases their confidence about doing well on the routine (self-efficacy) and enhances actual performance. Incremental theorists will not be influenced by brand promises. Incremental theorists derive a sense of self-efficacy through opportunities to learn and develop their skills. Thus, simply using a brand that promises they can perform well will not be effective in enhancing self-efficacy and task performance for incremental theorists. In contrast, incre- mental theorists engaged in a challenging workout routine can become more confident if they have engaged in oppor- tunities to learn proper workout technique. Thus, engaging in these self-development opportunities will enhance self- efficacy and task performance. Overview of Empirical Studies We examine our predictions in four studies. The first two studies demonstrate that brand use affects task performance for entity theorists using different brands (MIT and Gatorade), different tasks (GRE test and athletic exercise), and different ways of examining implicit self-theories (mea- sure vs. manipulation). Entity theorists performed better on difficult GRE questions (Study 1) and a physically challeng- ing athletic exercise (Study 2) when they used a brand promising better performance on the task; incremental theo- rists were unaffected by using these brands. In Study 3, we show that brand use affects entity and incre- mental theorists differently because they form self-efficacy in different ways. We compare two ways of increasing self- efficacy for athletic performance: using the Gatorade brand (tap water in a Gatorade cup) versus receiving training tips for better performance. Using the Gatorade brand increases self-efficacy and task performance only for entity theorists, whereas receiving training tips increases self-efficacy and task performance only for incremental theorists. We also find that self-efficacy mediates the relationship between implicit self-theories and task performance. In Study 4, we provide evidence that the beneficial effects of brand use are due to the brands promise of better performance. We manipulate the credibility of the Gatorade brand by providing information that either confirms or dis- confirms its promise of better athletic performance. Entity theorists experience a boost in self-efficacy and task per- formance only when the brand promise is confirmed, whereas incremental theorists are unaffected by whether the Gatorade brand promise is confirmed or disconfirmed. These findings support our theorizing that a brands promise of better performance is the source of self-efficacy (and bet- ter performance) for entity theorists. STUDY 1 Overview Undergraduate students took a GRE math test using either a pen engraved with the MIT name or a regular pen. Using the MIT pen resulted in better performance on the GRE test for entity theorists, particularly for more difficult GRE questions. Test performance for incremental theorists was unaffected by the pen they used. In addition, entity the- orists using the MIT pen performed better on the GRE test than did incremental theorists. Stimuli Selection We selected a GRE math test as the task for several rea- sons. First, the test could be constructed with difficult and easy questions, using data on the percentage of test takers answering questions correctly supplied by Educational Test- ing Services (which administers the GRE test). This feature enabled us to examine the idea that brands serve as a source of self-efficacy, which is required for difficult but not for easy tasks. Second, students had little to no experience with the GRE math test, which eliminated a potential confound with performance on the test. With this task in mind, we selected MIT as the brand for the study. As with most colleges and universities, MITs pri- mary mission is to increase the knowledge and skills of its student body, and this is the promise offered to students who attend. To confirm this notion, we asked a sample of under- graduate students (n = 43) to agree or disagree with four statements related to MITs brand promise, such as MIT promises to help students enhance their intellectual capabil- ities and MIT promises to help students succeed in chal- lenging intellectual tasks (a = .81). For comparison pur- poses, we asked students to respond to these same items for the control brand (Pilot) used in the study (a = .95). Partici- pants strongly agreed with the four statements about MIT, and the level of agreement was significantly stronger than for the Pilot brand (M MIT = 6.56 vs. M Pilot = 3.39; t(42) = 13.70, p < .001). Furthermore, these ratings did not differ for participants identified as entity versus incremental theo- rists (ps > .60). Finally, we selected a pen as the branded item. Partici- pants needed some type of writing instrument to take the test, so using a pen was unobtrusive. In addition, a pen lacks functional attributes that would help the user answer math questions more accurately, which is important for two rea- sons. First, it enables us to separate brand effects from the functional qualities of a product. Second, it enables us to rule out a placebo effect as an explanation for increases in test performance for students using the MIT pen. Placebo effects occur when people form false beliefs that an inert substance (e.g., a sugar pill) has functional qualities (e.g., active chemical ingredients) that yield positive results (e.g., cure illness). These false beliefs are triggered by the fact that pills typically do have active ingredients designed to cure illnesses. In our study, we use a product category (pens) that is not associated with functional qualities that improve peoples ability to solve difficult math problems. There is no reason to expect that a pen will improve peo- ples math abilities, and thus, better performance from using the MIT pen cannot be attributed to a placebo effect. Sample and Procedure Eighty undergraduate students (42 men and 38 women) majoring in business, economics, or engineering at two large public universities participated in exchange for course credit. Participants who did not complete the study (n = 2) or whose time data were lost (n = 2) were removed from the final sample. Participants were told that they were going to participate in several different studies to reduce suspicion that measures and procedures administered at different times were related to one another. The first study was a sur- vey, which included items measuring implicit self-theory, background questions, and demographics. For the second study, participants were told that the uni- versity bookstore was going to revamp its selection of pens and was asking for help in evaluating which pens people like most. Participants were randomly assigned to the MIT pen condition (n = 40) or the non-MIT pen condition (n = 36). In the MIT pen condition, participants were asked to select a pen from two options, a pen engraved with the MIT name or a pen from a less prestigious university. As we expected, all respondents chose the MIT pen. In the non- MIT pen condition, participants selected a pen from two options (a Pilot or Uni-Ball pen), and all choose the more appealing Pilot pen. Then, they were asked to use the pen while completing the next study and were told that they would be asked their opinions about the pen at the end of the next study. In the third study, participants were told that the univer- sity was interested in the readiness of undergraduate stu- dents for future graduate work. To measure readiness, they were asked to take a test consisting of 30 math questions from the GRE practice test booklet, using the pen they had selected earlier. The GRE test is a standardized test designed to measure academic abilities, and students were told that the test has been shown to correlate with intelli- gence tests. Participants had 30 minutes to complete the test and were told that they did not have to complete all the questions and could stop taking the test whenever they wanted. To enable participants to choose the level of diffi- culty they wanted to try, we phrased each test question to include information about the ease or difficulty of answer- ing the question (percentage of test takers answering the question correctly). Of the 30 questions, 15 were difficult (answered correctly by less than 60% of GRE test takers), and 15 were easy (answered correctly by more than 60% of GRE test takers). After the test, participants evaluated the pen they used for the test. Next, they were asked about the purpose of the study, and none of the participants answered correctly. Finally, participants were debriefed and thanked. In total, the study took approximately 50 minutes to complete. Measures GRE test performance. The number of difficult questions answered correctly (015) and the number of easy questions answered correctly (015) was recorded. We also measured the time spent taking the GRE test and included it as a con- trol variable in the main analysis. Implicit self-theory. We assessed belief in entity versus incremental theories of personality using the Implicit Per- 236 JOURNAL OF MARKETING RESEARCH, APRIL 2014 Brand Use, Self-Efficacy, and Performance 237 sons Theory Measure (Levy, Stroessner, and Dweck 1998). Participants responded to four statements representative of entity theory (e.g., Everyone is a certain kind of person, and there is not much that they can do to really change that) and four representative of incremental theory (e.g., Everyone, no matter who they are, can significantly change their basic characteristics), on a scale from 1 (strongly disagree) to 6 (strongly agree). We combined responses for all eight items into a scale (a = .90) after reversing responses for the entity theory items. Higher scores indicate a stronger belief in incremental theory. Note that scores for the implicit self-theory measure were not sig- nificantly correlated with self-reported grade point average (r = .17, p > .15), which could be associated with GRE test performance. Pen evaluation. After taking the GRE test, participants evaluated the pen (nice design and comfortable grip) on a scale from 1 (not at all) to 7 (very much). We com- bined responses to these two items (a = .72) and included them as a control measure in the main analysis to account for any unpleasant experiences using the pen. Results We used a mixed model analysis to test our prediction that entity theorists would perform better on the GRE test when using an MIT versus regular pen, but incremental the- orists would not. Furthermore, we expected to observe the beneficial effect of using the MIT pen for only the difficult GRE questions, on which test takers tend to struggle the most and for which the beneficial effects of using the MIT pen should be most pronounced. Our model included performance on the GRE test as the dependent measure, with implicit self-theory (continuous variable) and pen con- dition (non-MIT pen = 0, MIT pen = 1) as between-subjects variables and GRE question type (difficult = 0, easy = 1) as a within-subject variable. Following Singer (1998), we used MIXED in SPSS, which enables us to test effects for the between-subjects and within-subject variables simultane- ously. We centered scores for the implicit self-theory meas- ure to reduce potential issues stemming from multicollinear- ity (Aiken and West 1991) and included pen evaluation and time spent completing the GRE test as control variables. As we expected, the three-way interaction, implicit self- theory pen condition GRE question type, was signifi- cant (b = 1.14, t(72) = 2.01, p < .05), even after controlling for pen evaluations (b = .17, t(70) < 1, not significant [n.s.]) and time spent completing the test (b = .18, t(70) = 4.13, p < .01). In the following subsections, we examine this three-way interaction in more detail and report findings within each GRE question type (difficult vs. easy) sepa- rately to simplify presentation of results. Difficult GRE questions. Support for our prediction emerged in the form of a significant interaction between implicit self-theory and pen condition (b = 1.4, t(70.37) = 2.23, p < .05). Figure 1 illustrates the effect, which is plotted at one standard deviation below the mean (1 SD: entity theorists) and one standard deviation above the mean (+1 SD: incremental theorists) of the implicit self-theory meas- ure by substituting these values into the model (Cohen and Cohen 1983). To explore this interaction in more detail, we conducted simple slope tests at values one standard devia- tion above and below the mean of implicit self-theory (Aiken and West 1991; Cohen and Cohen 1983). We found a significant, positive relationship between pen condition (non-MIT pen = 0, MIT pen = 1) and performance on the GRE test for entity theorists (1 SD; b = 1.95, t(69.45) = 2.33, p < .05) but not for incremental theorists (+1 SD; b = .78, t(70.5) < 1, n.s.). As we predicted, only entity theorists performed better on the test when using the MIT pen versus a regular pen. Additional simple slope tests within each pen condition (West, Aiken, and Krull 1996) revealed that in the MIT pen condition, there was a significant, negative rela- tionship between implicit self-theories and performance, revealing that entity theorists performed better than incre- mental theorists (b = 1.22, t(69.75) = 2.81, p < .01). How- ever, in the non-MIT pen condition, there was no significant difference in performance between entity theorists and incremental theorists (b = .18, t(70.37) < 1, n.s.). Easy GRE questions. As we anticipated, the interaction between implicit self-theory and pen condition was not sig- nificant (b = .25, t(70.94) < 1, n.s.). Here, on items for which test takers do not struggle as much to answer the questions, the effect of using the MIT pen we found previ- ously with entity theorists was not evident. Discussion Our findings show that using an MIT pen increased test scores for some, but not all, students. We found that entity Figure 1 STUDY 1: PERFORMANCE ON THE GRE MATH TEST AS A FUNCTION OF IMPLICIT SELF-THEORY, BRAND USE CONDITION, AND QUESTION DIFFICULTY A: Difficult Questions B: Easy Questions 15 12 9 6 3 P e r f o r m a n c e Entity Theorists (1 SD) Incremental Theorists (+1 SD) MIT pen Regular pen 15 12 9 6 3 P e r f o r m a n c e Entity Theorists (1 SD) Incremental Theorists (+1 SD) MIT pen Regular pen theorists performed better on difficult GRE math questions when using an MIT versus regular pen, whereas the per- formance of incremental theorists was unaffected by the pen they used. These findings support our predictions about the greater influence of brands on task performance for entity versus incremental theorists. In the next study, we switch to a different context to repli- cate our findings. First, we examine a different task domain, strenuous athletic exercise. Second, we use a task in which the level of difficulty increases as performance increases. We asked participants to press a handgrip, which is initially easy to press but becomes more difficult (due to fatigue) as the handgrip is pressed more times. Third, we used a differ- ent brand, Gatorade, which offers a brand promise (better athletic performance) related to the task domain. We also manipulate implicit self-theories in the next study. Because we measured implicit self-theory as an indi- vidual difference factor in Study 1, one might question whether preexisting differences between entity and incre- mental theorists could have contributed to test performance differences. Although researchers usually measure implicit self-theory, others have shown that these beliefs can be manipulated by exposing people to information advocating a particular theory (Chiu, Hong, and Dweck 1997; Park and John 2010; Yorkston, Nunes, and Matta 2010). We follow their approach and expose participants to an article present- ing scientific evidence either that personal qualities are enduring and cannot be easily changed (to manipulate entity theory) or that personal qualities are malleable and can be developed (to manipulate incremental theory). STUDY 2 Sample and Procedure One hundred seven undergraduate students (62 women and 45 men) participated in a 2 (implicit self-theory manipulation: entity, incremental) 2 (brand use: Gatorade, Ice Mountain) between-subjects experiment in exchange for extra credit. Participants who had problems with the handgrip counter (n = 4) were removed from the final sample. Participants were told they were going to participate in several different studies. First, participants completed a survey that included per- sonal background and demographic information. Next, a second survey was administered that contained the implicit self-theory manipulation (described subsequently). Finally, in the third study, participants were told that Gatorade (Ice Mountain) was interested in consumer opinions about a new bottled water it was planning to introduce. In the Gatorade (Ice Mountain) condition, participants were given a Gatorade paper cup (plain paper cup) filled with tap water. All participants were asked to sign a form if they agreed to evaluate the water, which was intended to discourage them from discounting their brand experience as something they were forced to do, thus potentially decreasing feelings of self-efficacy (for a similar approach, see Jones et al. 1981). Next, participants were told that prior research had found that consumers make more accurate evaluations about a new product when they experience the product repeatedly. Therefore, they were asked to hold the cup and continue drinking water from it while they completed several tasks, including copying line figures, circling a vowel in para- graphs, and exercising with a handgrip. The first two tasks were included as filler tasks to reduce suspicion about the handgrip task. Participants were given 15 minutes to exer- cise with the handgrip, but no specific goal was set (for a similar approach, see Muraven, Tice, and Baumeister 1998). After finishing the handgrip task, participants evaluated the water they drank during the exercise and were asked about the purpose of the experiment (which none correctly guessed). Finally, participants were debriefed and thanked. In total, the study took 30 minutes to complete. Stimuli Selection We asked participants to press a handgrip, which had a counter to record the number of times the handgrip was pressed. Although easy to press initially, the handgrip was much more difficult to squeeze as it was pressed more times. Similar handgrip tasks have been used in self-regulation research (see Muraven, Tice, and Baumeister 1998). We selected Gatorade as the focal brand for the study. Gatorades brand promise is to enhance athletic performance, which fits well with the handgrip task. To confirm awareness of this promise, we asked undergraduate students (n = 42) to agree or disagree with four statements about Gatorades brand promise, such as Gatorade promises to help con- sumers enhance their athletic capabilities and Gatorade promises to help consumers succeed in strenuous athletic exercise (a = .92). For comparison, students responded to these same items for the other brand of water (Ice Moun- tain) used as a control brand (a = .93). Participants strongly agreed with the four statements about Gatorade, and the level of agreement was significantly stronger than for Ice Mountain (M Gatorade = 6.38 vs. M Ice Mountain = 4.52; t(41) = 9.67, p < .001). Furthermore, these ratings did not differ for participants in the entity versus incremental theory condi- tion (ps > .20). Finally, we chose bottled water as the focal product. Bot- tled water lacks functional attributes that would help users increase their handgrip strength, which is important for two reasons. First, it enables us to separate brand effects from the functional qualities of the product that participants con- sumed. Second, as described previously, the absence of functional qualities also differentiates our experimental con- text from placebo effect studies. For example, in placebo studies using energy drink brands (Irmak, Block, and Fitzsi- mons 2005), participants have been told that the beverage is an energy drink, which facilitates the formation of false beliefs about the functional qualities of the drink. In our study, participants were told they were drinking bottled water, and subsequent testing confirmed that they perceived it as regular water. Implicit Self-Theory Manipulation We manipulated implicit self-theories by having partici- pants read an article presenting views consistent with entity or incremental theory (Chiu, Hong, and Dweck 1997). We told participants that we were interested in their opinions about the articles. In addition, we asked them to write a short essay supporting the authors viewpoint. The follow- ing is a sample from each article: In his talk at the American Psychological Associations annual convention held at Washington D.C. in August, Dr. George Medin argued that in most of us, by the age 238 JOURNAL OF MARKETING RESEARCH, APRIL 2014 Brand Use, Self-Efficacy, and Performance 239 of ten, our character has set like plaster and will never soften again. He reported numerous large longitudinal studies showing that people age and develop, but they do so on the foundation of enduring dispositions. (Entity Theory) In his talk at the American Psychological Associations annual convention held at Washington D.C. in August, Dr. George Medin argued that no ones character is as hard as a rock so that it cannot be changed. Only for some, greater effort and determination are needed to effect changes. He reported numerous large longitudi- nal studies showing that people can mature and change their character. He also reported research findings showing that peoples personality characteristics can change, even in their late sixties. (Incremental Theory) We conducted a pilot test of the manipulation with under- graduate students (n = 56). Participants read either the entity theory article or the incremental theory article and then completed a survey that included a measure of implicit self- theory beliefs (Implicit Persons Theory Measure described in Study 1; a = .95). The results confirmed that participants who read the entity theory article were more likely to agree with entity theory beliefs than those who read the incremen- tal theory article (t(54) = 1.85, p < .05). Measures Handgrip performance. We measured performance by the number of times the handgrip was pressed, which was auto- matically recorded by a counter incorporated into the hand- grip. We also measured the total time participants spent pressing the handgrip and included it as a control variable in the main analysis. Water evaluation. Participants evaluated the water they drank on two attributes (crisp and fresh) on a 1 (not at all) to 7 (very much) scale. We combined responses to these two items (a = .87) and included the resulting meas- ure as a control variable in the main analysis to account for any unpleasant experiences drinking the water. Results We predicted that participants in the entity theory condi- tion would perform better when drinking Gatorade versus Ice Mountain but that participants in the incremental theory condition would be unaffected by the brand of water con- sumed. We performed a 2 (implicit self-theory manipula- tion: entity, incremental) 2 (brand use: Gatorade, Ice Mountain) analysis of covariance on handgrip performance, with water evaluation and time spent on the handgrip task as covariates. As we expected, a significant interaction between implicit self-theory and brand use condition emerged (F(1, 97) = 3.82, p < .05; see Figure 2), even after controlling for water evaluation (F(1, 97) < 1, n.s.) and time spent on the handgrip task (F(1, 97) = 102.29, p < .01). Planned comparisons showed that participants in the entity theory condition performed better on the handgrip task when drinking Gatorade water (tap water in a Gatorade cup) than Ice Mountain water (tap water in a plain cup) (F(1, 97) = 6.92, p < .05). However, participants in the incremental theory condition were not affected by the brand of water consumed (F(1, 97) < 1, n.s.). Additional planned comparisons showed that, among participants who drank Gatorade water, those in the entity theory condition per- formed better on the handgrip task than those in the incre- mental theory condition (F(1, 97) = 5.69, p < .05). However, we did not find such a difference among participants drink- ing Ice Mountain water (F(1, 97) < 1, n.s.). Finally, we examined several background factors. Partici- pants were randomly assigned to the entity versus incre- mental theory conditions, which should rule out individual differences that could affect performance. We did, however, ask participants about their attitudes toward Gatorade (1 = do not like it at all, and 7 = like it very much) and exer- cise frequency (1 = every day, and 7 = once very year) at the beginning of the study. We found no significant differ- ences for participants in the entity and incremental theory conditions on these factors (ps > .60). Discussion Our findings replicate results from the first study with a different brand, different brand promise, and different task. Furthermore, our results, obtained using a manipulation of implicit self-theory, replicate findings from Study 1, in which we measured implicit self-theory as an individual dif- ference variable. Participants who read an article espousing entity theory beliefs performed the handgrip task better when consuming Gatorade versus Ice Mountain water, whereas participants in the incremental theory condition were unaffected by the brand of water consumed. These findings support our prediction that entity theorists perform better when using a brand that promises better per- formance on the task. We did wonder, however, whether we could obtain the same effect if participants were simply exposed to the brand (priming) rather than using it during the handgrip task. To explore this possibility, we ran an additional condition in which participants (n = 65) were exposed to the Gatorade cup (which was placed on their desk with some other study materials) but actually drank water from a cup with a fictitious brand name (HydroPhase) not associated with any particular benefit or promise related to athletic performance in our pretests. If priming is a viable mechanism, we should observe better handgrip performance among entity (vs. incremental) theorists. However, there was no significant difference in handgrip performance 180 160 140 120 100 80 60 40 20 0 P e r f o r m a n c e Entity Theory Manipulation Incremental Theory Manipulation Figure 2 STUDY 2: PERFORMANCE ON THE HANDGRIP TASK AS A FUNCTION OF IMPLICIT SELF-THEORY MANIPULATION AND BRAND USE CONDITION Gatorade Ice Mountain between entity and incremental theorists (F(1, 61) = 1.24, p > .25). Thus, simply being exposed to the brand (brand priming) did not increase handgrip performance for entity theorists. We also entertained the possibility that mood effects might be at play. Perhaps using brands makes entity theo- rists (vs. incremental theorists) feel more positive, and this facilitates better performance on the handgrip task. To examine this possibility, we ran a condition in which partici- pants (n = 58) were asked to drink water in a cup with the Gatorade name or a fictitious brand name (HydroPhase). The procedure followed that of Study 2, except that partici- pants completed the Positive and Negative Affect Schedule mood measure (Watson, Clark, and Tellegen 1988) and the implicit self-theory measure. We conducted a multiple regression analysis with positive mood as the dependent measure and implicit self-theory, brand use condition, and their interaction as independent variables. We conducted the same analysis on negative mood. Results of both analyses revealed no significant main effects or interactions (ps > .20), confirming that mood is not the factor responsible for better performance among entity theorists. In the next study, we delve deeper into the process. First, we measure self-efficacy and show that using the Gatorade brand increases self-efficacy (and handgrip performance) for only entity theorists. Second, we show why incremental theorists are not affected by brand use. Previously, we rea- soned that incremental theorists form self-efficacy by engaging in opportunities to learn and develop their skills in a domain. In our prior studies, no such opportunity was pro- vided, and thus, their handgrip performance did not show improvement. In Study 3, we provide a learning opportunity by offering training tips to improve skills on the handgrip task and find that self-efficacy increases (and handgrip per- formance increases) for incremental theorists. As we expected, training tips do not affect entity theorists. Overall, we show that self-efficacy mediates performance on the handgrip task. We find that handgrip performance is better for entity theorists who drink Gatorade water versus receiving training tips, and this difference is mediated by self-efficacy. Similarly, we show that handgrip performance is better for incremental theorists who receive training tips versus drinking Gatorade water, and this difference is also mediated by self-efficacy. STUDY 3 Sample and Procedure One hundred nine undergraduate students (37 women and 72 men) participated for extra course credit and were ran- domly assigned to one of the experimental conditions: (1) Gatorade (n = 38), (2) HydroPure (control brand) (n = 34), and (3) HydroPure + training tips (n = 37). Participants who were unable to press the handgrip (n = 1) or had problems with the handgrip counter (n = 1) were removed from the final sample. As in our prior studies, participants were told that they were going to participate in several different studies. First, participants completed a survey, with the Implicit Person Theory Measure embedded among other questions. Next, participants in the Gatorade (HydroPure) condition were told that Gatorade (a national manufacturer) was interested in consumer opinions about a new bottled water the company was planning to introduce and were asked to drink water in a cup with the Gatorade (HydroPure) name while performing several tasks, including the handgrip task, using the same procedure as Study 2. Participants in the HydroPure + train- ing tips condition read a one-page article offering tips useful for the handgrip exercise before doing the handgrip task. After finishing the handgrip task, participants completed the self-efficacy measure (described subsequently) and evalu- ated the water they had consumed. Then, participants were asked about the purpose of the study, and none correctly guessed the purpose. Finally, participants were debriefed and thanked. The study took 30 minutes to complete. Stimuli Brands. We used HydroPure, a fictitious brand of bottled water, as the control brand for the study. To check percep- tions of this brand, we asked undergraduate students (n = 42) to agree or disagree with four statements about HydroPure, such as HydroPure promises to help consumers enhance their athletic capabilities and HydroPure promises to help consumers succeed in strenuous athletic exercise (a = .91). Comparing their responses with the same statements for Gatorade, we confirmed that participants rated Gatorade higher than HydroPure in terms of promising better athletic performance (M Gatorade = 6.38 vs. M HydroPure = 4.48; t(41) = 11.67, p < .001). Furthermore, these ratings did not differ for entity versus incremental theorists (p > .20). Training tips. We developed an article with tips for several types of exercise, including handgrips. The article offered the following tip for handgrips: (1) press the grip for as long as comfortable (2) take a very short break for 23 seconds, and (3) try another round of hand-gripping with gradually increased speed. This tip is helpful because pressing the handgrip too rapidly in the beginning can result in get- ting tired too quickly and giving up pressing the handgrip. We asked undergraduate students (n = 56) to read the article and rate it on several dimensions (bad/good, unfavor- able/ favorable, negative/ positive, unreliable/ reliable, untrustworthy/ trustworthy, incredible/ credible, unin- formative/ informative, not at all useful/ useful) on seven- point scales. We combined the responses to these items (a = .90), and the mean of the combined items (M = 4.79) was greater than the scale midpoint (t(55) = 9.02, p < .001), indi- cating that participants positively evaluated the information in the article. Moreover, these ratings did not differ for entity versus incremental theorists (p > .70). Pilot test. To confirm that entity (incremental) theorists view using the Gatorade brand (training tips) as a source of self-efficacy, we conducted a pilot study with undergraduate students (n = 56). We used the same procedure as in the main study but interrupted participants several minutes into the handgrip task to ask them questions. In the Gatorade condition, participants were asked to rate their agreement on the following items: Thinking about how Gatorade affects peoples athletic capabilities will make me more confident about doing better in the handgrip exercise, and Thinking about how Gatorade affects peoples athletic capabilities will make me feel more confident in my own abilities during the handgrip exercise (0 = strongly disagree, and 100 = strongly agree; a = .99). As we expected, a regression analysis revealed that entity theorists (vs. incremental theo- rists) were more likely to agree with these statements (b = 240 JOURNAL OF MARKETING RESEARCH, APRIL 2014 Brand Use, Self-Efficacy, and Performance 241 .35, t(25) = 2.11, p < .05), even after controlling for water evaluation (b = .41, t(25) = 2.47, p < .05). In the HydroPure + training tips condition, participants agreed or disagreed (using the same 100-point scale) with the following items: Thinking about how the training tips affect peoples athletic capabilities will make me more con- fident about doing better in the hand-grip exercise, and Thinking about how the training tips affects peoples ath- letic capabilities will make me feel more confident in my own abilities during the hand-grip exercise (a = .95). As we anticipated, incremental theorists (vs. entity theorists) were more likely to agree with these statements (b = .39, t(25) = 2.11, p < .05), even after controlling for water evaluation (b = .11, t(25) < 1, n.s.). These findings confirm that incremental theorists viewed the training tips as a source of self-efficacy, whereas entity theorists viewed the Gatorade brand as a source of self-efficacy. Measures Implicit self-theory. We used the same Implicit Persons Theory Measure (Levy, Stroessner, and Dweck 1998) as in Study 1. We combined responses for all items into a scale (a = .91). Handgrip performance. We measured performance by the number of times the handgrip was pressed. We also meas- ured the total time participants spent pressing the handgrip and included it as a control variable in the main analysis. Self-efficacy. After the handgrip task, participants rated how confident they were during the exercise with the state- ments I was confident that I would do well in the exercise, and I was confident in my ability to perform well in the exercise (1 = strongly disagree, and 7 = strongly agree; a = .90). Note that this measure is task specific, as opposed to measuring a general sense of self-efficacy, in line with guidelines for self-efficacy research (Bandura 2006). Water evaluation. Participants evaluated the water using the same scale described in Study 2 (a = .78). We included this evaluation as a control variable in the main analysis. Results We conducted a multiple regression analysis to test our predictions. The analysis included performance on the handgrip task as the dependent measure, with implicit self- theory (continuous variable), experimental condition (two dummy variables to represent the three conditions), and their interaction as the independent variables. As in the pre- vious studies, we centered scores for the implicit self-theory measure and included water evaluation and time spent on the handgrip task as control variables. As we expected, two significant interactions emerged (see Figure 3). One was a significant interaction between implicit self-theory and the first dummy variable (Gatorade = 0, HydroPure = 1) (b = .21, t(99) = 1.98, p = .05). The second was a significant interaction between implicit self-theory and the second dummy variable (Gatorade = 0, HydroPure + training tips = 1) (b = .38, t(99) = 3.75, p < .001). Next, we examine these interactions in more detail and report find- ings for entity and incremental theorists separately to sim- plify presentation of results. Incremental theorists. As we expected, we found that incremental theorists performed best on the handgrip task when they received training tips before the task. First, incre- mental theorists performed better when they received train- ing tips versus drinking Gatorade water, confirmed by a sig- nificant positive relationship between the dummy variable (Gatorade = 0, HydroPure + training tips = 1) and handgrip performance (b = .26, t(99) = 2.02, p < .05). Second, incre- mental theorists who drank HydroPure water performed better when they received training tips than when they did not, as shown by a significant positive relationship between the dummy variable (HydroPure = 0, HydroPure + training tips = 1) and handgrip performance (b = .37, t(99) = 2.69, p < .01). Also of note, and consistent with Study 2, incre- mental theorists did not perform better when drinking Gatorade versus HydroPure water, as confirmed by a non- significant relationship between the dummy variable (Gatorade = 0, HydroPure = 1) and handgrip performance (b = .11, t(99) = 1.03, p > .30). Entity theorists. As we expected, entity theorists per- formed best on the handgrip task when drinking Gatorade water. First, entity theorists performed better in the Gatorade versus HydroPure condition, confirmed by a sig- nificant negative relationship between the dummy variable (Gatorade = 0, HydroPure = 1) and handgrip performance (b = .44, t(99) = 3.44, p < .01). Second, entity theorists performed better when drinking Gatorade versus receiving training tips before the task, as shown by a significant nega- tive relationship between the dummy variable (Gatorade = 0, HydroPure + training tips = 1) and handgrip performance (b = .43, t(99) = 3.34, p < .01). Mediation analysis. We conducted a mediation analysis to test whether performance differences between entity and incremental theorists were mediated by feelings of self- efficacy. We expected entity theorists to feel a heightened sense of self-efficacy when using the Gatorade brand (vs. learning the training tips), thus resulting in better performance. In contrast, we expected incremental theorists to feel a heightened sense of self-efficacy when learning the training tips (vs. using the Gatorade brand), thus resulting in better performance. 120 100 80 60 40 20 0 P e r f o r m a n c e Entity Theorists (1 SD) Incremental Theorists (+1 SD) Gatorade HydroPure HydroPure use+ training tips Figure 3 STUDY 3: PERFORMANCE ON THE HANDGRIP TASK AS A FUNCTION OF IMPLICIT SELF-THEORY AND SELF-EFFICACY APPROACH CONDITION For this analysis, we focused on the Gatorade condition and the HydroPure + training tips condition. A bootstrap analysis using the INDIRECT SPSS macro (Preacher and Hayes 2008) confirmed a significant mediating pathway from the implicit self-theories condition interaction to per- formance on the handgrip task through self-efficacy (95% confidence interval [CI]: 5.82, 39.47). As we predicted, these results indicate that entity and incremental theorists rely on different approaches to enhance self-efficacy, which affects their performance on the handgrip task. Discussion In this study, we provide evidence that entity and incremen- tal theorists derive feelings of self-efficacy through different routes. Entity theorists enhance feelings of self-efficacy by using the Gatorade brand, which promises to help con- sumers achieve better athletic performance. Incremental theorists enhance feelings of self-efficacy by receiving training tips, which promise to improve athletic perform- ance. These differences in self-efficacy drive better task per- formance, as the mediation analyses show. In the next study, we provide further evidence pertaining to the source of self-efficacy for entity theorists. Previously, we reasoned that entity theorists use brands as a source of self-efficacy because brands offer promises that consumers can perform better. Thus far, we have shown that brand use increases self-efficacy and task performance; however, we have not isolated the effect to brand promises in particular. To do so, we manipulate the credibility of the Gatorade brand promise in Study 4. We asked participants to drink Gatorade water while performing the handgrip task but manipulated the credibility of the Gatorade brand promise (improved athletic performance) before the task. Participants read an article that summarized scientific evidence and expert opinion concluding that (1) Gatorade enhances athletic per- formance (promise confirmation) or (2) Gatorade does not enhance athletic performance (promise disconfirmation). We posit that entity theorists should benefit from using the Gatorade brand for the handgrip task only in the confirma- tion condition. In the disconfirmation condition, there is no longer an assurance that using the Gatorade brand will result in better athletic performance, which will result in entity theorists no longer viewing the brand as a valuable source of self-efficacy. In contrast, we expect incremental theorists to remain unaffected by this manipulation because they do not rely on brand promises as a source of self-efficacy. Through this manipulation, we also shed light on the con- scious versus nonconscious nature of the process whereby brand use affects task performance for entity theorists. In the confirmation condition, for example, we make the posi- tive link between the Gatorade brand and better athletic per- formance salient to participants. If brand use affects task performance nonconsciously, performance should not be enhanced when entity theorists are consciously aware of the positive effects of the Gatorade brand on athletic behavior (Kramer and Block 2008; Shiv, Carmon, and Ariely 2005). If brand use affects task performance consciously, however, performance should be enhanced when entity theorists are consciously aware of the positive effects of the Gatorade brand on athletic behavior. STUDY 4 Sample and Procedure One hundred fifty-one undergraduate students (88 women and 63 men) participated for extra course credit. They were randomly assigned to one of the experimental conditions: (1) Gatorade promise disconfirmation + Gatorade use (n = 47), (2) Gatorade promise confirmation + Gatorade use (n = 49), or (3) Gatorade promise confirmation + HydroPure (control brand) use (n = 55). Participants who were unable to press the handgrip (n = 1), did not follow instructions (n = 2), or ran out of time to complete the study (n = 6) were removed from the final sample. As in previous studies, participants were told they were going to participate in several different studies. First, participants completed a survey that included back- ground and demographic questions. Embedded in the sur- vey was the Implicit Self Theory Measure. Next, a second survey was administered that asked participants to read and give opinions about several articles on various consumer topics, including an article about Gatorade that either con- firmed or disconfirmed the brands promise of enhancing athletic performance. They were then asked to drink water from a Gatorade or HydroPure cup while performing sev- eral tasks, including the handgrip task. The same procedure from prior studies was used, with one exception. We pro- vided a difficult goal for the handgrip task (The top 2% of college students can press the handgrip for 600 times or more in 15 minutes) and instructed participants to try their best, enabling us to examine whether our prior results are robust when a challenging goal is made explicit before the task. After finishing the handgrip task, participants completed the same measures as in Study 3. Next, participants were asked about the purpose of the study, and none correctly guessed the purpose. Finally, participants were debriefed and thanked. In total, the study took 30 minutes. Brand Promise Manipulation We developed two versions of the Gatorade article. For the promise confirmation condition, the article cited evi- dence that Gatorade helps consumers enhance athletic per- formance (e.g., The Texas Medical Association reports that Gatorade can reduce elevated heart rates in athletes, which allows them to continue their activities) and concluded with a statement confirming Gatorades promise (Can Gatorade make you more athletic? The resounding answer is YES). For the promise disconfirmation condition, the article cited evidence that Gatorade does not help con- sumers enhance athletic performance (e.g., The Texas Medical Association reports that Gatorade can cause ele- vated heart rates in athletes, which causes them to cut back on their activities) and concluded with a statement discon- firming Gatorades promise (Can Gatorade make you more athletic? The resounding answer is NO). We conducted a pilot test of the manipulation with under- graduate students (n = 58) similar to those in the main study. Participants read one of the Gatorade articles and were then asked to agree or disagree (1 = strongly agree, and 7 = strongly disagree) with eight statements (e.g., I think that the article confirms Gatorades promise of better athletic performance). We combined responses to these items (a = 242 JOURNAL OF MARKETING RESEARCH, APRIL 2014 Brand Use, Self-Efficacy, and Performance 243 .98), and as we expected, participants agreed more with these statements after reading the confirmation versus dis- confirmation article (M confirmation = 2.84 vs. M disconfirmation = 5.78; t(56) = 10.18, p < .001). Furthermore, we conducted a regression analysis using responses to these items as the dependent variable, including brand promise manipulation condition, implicit self-theories (a = .92), and their inter- action as independent variables. The results confirm that entity and incremental theorists did not differ in their agree- ment with the statements after reading the confirmation ver- sus disconfirmation article, as indicated by the lack of a sig- nificant interaction effect (b = .04, t(54) < 1, n.s.). Measures We used the same measures as in Study 3: implicit self- theory (a = .87), handgrip performance, self-efficacy (a = .98), and water evaluation (a = .87). We also asked partici- pants to agree or disagree with two statements (a = .88) comparing the water they drank with regular water (e.g., The Gatorade water I drank today is similar to regular bot- tled water) to examine whether the confirmation (discon- firmation) article influenced perceptions of the water as having (not having) unique ingredients to enhance athletic performance. Results We conducted a multiple regression analysis to test the prediction that entity theorists would perform better when drinking Gatorade water if they read the article confirming Gatorades brand promise but not if they read the article dis- confirming Gatorades brand promise. Incremental theo- rists, who do not rely on brand promises as a source of self- efficacy, would not be affected by the brand promise manipulation. The analysis included performance on the handgrip task as the dependent measure, with implicit self- theory (continuous variable), experimental condition (two dummy variables to represent the three conditions), and the interaction between implicit self-theory and experimental condition as the independent variables. We centered scores for the implicit self-theory measure and included water evaluation and time spent on the handgrip task as control variables. As we expected, there was a significant interaction between implicit self-theory and the first dummy variable (promise disconfirmation + Gatorade use = 0, promise con- firmation + Gatorade use = 1) (b = .14, t(134) = 2.08, p < .05; see Figure 4). This finding confirmed that the brand promise manipulation affected entity and incremental theo- rists differently. In contrast, the interaction between implicit self-theory and the second dummy variable (promise dis- confirmation + Gatorade use = 0, promise confirmation + HydroPure use = 1) was not significant (b = .01, t(134) < 1, n.s.). This result shows that when the Gatorade brand promise was disconfirmed, using the Gatorade brand did not affect entity and incremental theorists differently; indeed, using the Gatorade brand (vs. HydroPure) did not result in better performance for entity theorists. In the following sub- sections, we examine the interactions in more detail and report findings for entity and incremental theorists sepa- rately to simplify presentation of results. Entity theorists. As we predicted, entity theorists per- formed better on the handgrip task when using the Gatorade brand and reading the article confirming (vs. disconfirming) Gatorades brand promise, as shown by a significant posi- tive relationship between the dummy variable (promise dis- confirmation + Gatorade use = 0, promise confirmation + Gatorade use = 1) and performance on the handgrip task (b = .32, t(134) = 3.65, p < .001). Furthermore, reading the arti- cle confirming Gatorades brand promise, without using the Gatorade brand, did not enhance performance on the hand- grip task; that is, brand use was required. We observe evi- dence to this effect in the significant negative relationship between the dummy variable (promise confirmation + Gatorade use = 0, promise confirmation + HydroPure use = 1) and performance on the handgrip task (b = .20, t(134) = 2.22, p < .05). Finally, reading the article disconfirming Gatorades brand promise did not harm performance among entity theorists using the Gatorade brand, as shown by a nonsignificant relationship between the dummy variable (promise disconfirmation + Gatorade use = 0, promise con- firmation+ HydroPure use = 1) and performance on the handgrip task (b = .13, t(134) = 1.43, p > .15). Incremental theorists. Incremental theorists using the Gatorade brand were not affected by the brand promise manipulation, indicated by a nonsignificant relationship between the dummy variable (promise disconfirmation + Gatorade use = 0, promise confirmation + Gatorade use = 1) and performance on the handgrip task (b = .06, t(134) < 1, n.s.). In addition, consistent with prior studies, drinking Gatorade (vs. HydroPure) water did not improve handgrip performance for incremental theorists, even when brand use was preceded by the article confirming Gatorades promise of better athletic performance. This result is indicated by a nonsignificant relationship between the dummy variable (promise confirmation + Gatorade use = 0, promise confir- mation + HydroPure use = 1) and handgrip performance (b = .05, t(134) < 1, n.s.). Mediation analysis. We conducted a meditation analysis to test whether the effect of the brand promise manipulation on performance was mediated by feelings of self-efficacy. 120 100 80 60 40 20 0 P e r f o r m a n c e Entity Theorists (1 SD) Incremental Theorists (+1 SD) Gatorade promise disconfirmation + Gatorade use Gatorade promise confirmation + Gatorade use Gatorade promise confirmation + HydroPure use Figure 4 STUDY 4: PERFORMANCE ON THE HANDGRIP TASK AS A FUNCTION OF IMPLICIT SELF-THEORY AND EXPERIMENTAL CONDITION Entity theorists should feel a heightened sense of self-efficacy when using the Gatorade brand in the brand promise confir- mation (vs. disconfirmation) condition, thus resulting in better performance. In contrast, incremental theorists should not be affected by the brand promise manipulation. To test this mediating process, we focused on participants who used the Gatorade brand: (1) those in the Gatorade promise disconfirmation + Gatorade use condition and (2) those in the Gatorade promise confirmation + Gatorade use condition. A bootstrap analysis using the INDIRECT SPSS macro (Preacher and Hayes 2008) confirmed a significant mediating pathway from the interaction between implicit self-theories and experimental condition to performance on the handgrip task through self-efficacy (95% CI: 19.96, .06). Thus, self-efficacy mediates the interaction effect of implicit self-theory and experimental condition on task per- formance. Entity theorists are more likely to rely on brand use to enhance feelings of self-efficacy if the brand promise is confirmed (vs. disconfirmed), which results in better hand- grip performance. In contrast, incremental theorists, who do not use brands to enhance self-efficacy, are not affected by whether the brand promise is confirmed or disconfirmed. Supplementary findings. To rule out an alternative expla- nation for our results, we examined whether perceptions of the Gatorade water were equivalent for the brand promise confirmation and disconfirmation conditions. Although par- ticipants drank regular water, those in the promise confir- mation (vs. disconfirmation) condition may have thought the Gatorade water was produced to enhance athletic per- formance. We asked participants to rate how similar the Gatorade water was to regular bottled water, using the measure described previously. We found no significant dif- ference in these evaluations for the promise confirmation versus disconfirmation condition (M confirmation = 81.57 vs. M discomfirmation = 75.30; t(89) = 1.24, p > .20). Furthermore, entity and incremental theorists did not differ in their per- ceptions of the water, as indicated by a nonsignificant inter- action between implicit self-theory and brand promise con- dition (b = .05, t(86) < 1, n.s.). Thus, even though entity theorists viewed Gatorade water as similar to regular water, using the Gatorade brand enhanced their self-efficacy as long as its brand promise was confirmed. Discussion We found further evidence that entity theorists use brand promises as a source of self-efficacy, which results in better task performance. Entity theorists benefited from using the Gatorade brand when the brand promise was confirmed but not when it was disconfirmed. When the brand promise was confirmed, entity theorists used it to assure themselves of better athletic performance. However, when the brand promise was disconfirmed, it could no longer provide a sense of assurance, and therefore, using the Gatorade brand failed to boost self-efficacy (and athletic performance) for entity theorists. Note that self-efficacy and performance did not decline for entity theorists when the brand promise was disconfirmed. Consistent with our theorizing, this finding suggests that entity theorists seek out brands as a source of self-efficacy when facing challenging tasks, and when a brand no longer helps or can hurt their feelings of self-efficacy, they stop relying on the brand as a source of self-efficacy to drive better performance. These findings also suggest that brand use affects task performance for entity theorists through a conscious process. Our manipulations of Gatorades brand promise made the effect of the brand on athletic performance salient to participants. If brand use affects task performance on a nonconscious basis, performance should not be enhanced when entity theorists are consciously aware of the Gatorade brands positive effects on athletic behavior (Kramer and Block 2008; Shiv, Carmon, and Ariely 2005). Yet we found that entity theorists who used the Gatorade brand after read- ing that Gatorade enhances performance (confirmation con- dition) improved their handgrip performance. Furthermore, entity theorists who used the Gatorade brand after reading that Gatorade harms performance (disconfirmation condi- tion) did not perform worse than entity theorists who read the confirmation article but used the HydroPure brand. If the Gatorade article unconsciously affected entity theorists, they should have performed worse in the disconfirmation condition. Thus, taken together, these results support the idea that entity theorists consciously benefit from brand use. GENERAL DISCUSSION When a person struggles with a difficult task, can brand use increase his or her level of performance? The answer is yes. Across four studies, we found that consumers with cer- tain implicit self-theories (entity theorists) benefit from using brands that promise better performance. Using a brand such as Gatorade increased their sense of self-efficacy in performing a challenging athletic task, and as a result, entity theorists increased their performance in the task. In contrast, incremental theorists did not benefit from brand usage, either in terms of self-efficacy or task performance. Further results supported the view that self-efficacy mediates the relationship between brand usage and task performance and that the process is conscious in nature. Contributions to Branding Research Our findings add to an increasing body of research show- ing that brands deliver self-related benefits. Consumers use brands to express and enhance their self-images, and using these brands can actually enhance self-perceptions about their personality traits (Park and John 2010). Beyond these purely perceptual effects, we show that using brands can also enhance self-efficacy, which is an important regulatory mechanism that governs the level of challenge people are willing to undertake, resulting in better task performance. To illustrate the difference between self-perceptions and self-efficacy, consider research showing that using a brand with a strong personality (Victorias Secret) can enhance consumers self-perceptions about brand-related personality traits (feminine and glamorous), if they are entity theorists (Park and John 2010). Could this self-perception effect be the key determinant of better handgrip performance for entity theorists drinking Gatorade water? We explored this question by measuring self-perceptions of athleticism (along with self-efficacy) in Study 3 and included both measures in a mediation analysis similar to the one described in Study 3. When we included self-perceptions of athleticism and self-efficacy as mediating factors simultane- ously, a bootstrap analysis using the INDIRECT SPSS macro (Preacher and Hayes 2008) confirmed self-efficacy as the mediating factor (95% CI: 3.25, 34.89) but not self- 244 JOURNAL OF MARKETING RESEARCH, APRIL 2014 Brand Use, Self-Efficacy, and Performance 245 perceptions (95% CI: 1.96, 18.21). Thus, although self- perceptions may have changed by drinking Gatorade water, these self-perceptions were not as effective as self-efficacy in driving the improvement in handgrip performance for entity theorists. Our results also complement prior research on noncon- scious brand priming, which shows that incidental exposure to brands can automatically elicit goals and behaviors (Chartrand et al. 2008; Fitzsimons, Chartrand, and Fitzsi- mons 2008). Here, we show that brands can affect behavior through a conscious route. In Study 2, we ruled out the pos- sibility that our results were due to a nonconscious process by running a separate condition in which participants were exposed to the Gatorade cup but did not drink water from it. Merely being exposed to the Gatorade name/logo did not enhance task performance for entity theorists. Additional evidence emerged in Study 4, in which entity theorists who drank water from a Gatorade cup increased their handgrip performance even when they were made consciously aware that the Gatorade brand positively affects athletic perform- ance. If Gatorade affected handgrip performance through a nonconscious process, performance should not have been enhanced when the link between Gatorade and athletic per- formance was made salient (Kramer and Block 2008; Shiv, Carmon, and Ariely 2005). Moreover, our findings provide evidence that self-efficacy beliefs, which are formed through a deliberative process (Gist and Mitchell 1992), mediate the effect of brand use on behavior (task performance). In Studies 3 and 4, we provide evidence of mediation using a measure of self-efficacy administered after completion of the handgrip task, which avoided the problem that administering the measure before or during the task could affect subsequent task performance (Feldman and Lynch 1988). To provide further evidence that (1) brand use increases self-efficacy (for entity theo- rists) and (2) increases in self-efficacy are accompanied by greater effort being expended on the task, we conducted a small study in which self-efficacy and effort were measured before the task. To provide a sense of the task, we gave undergraduate students (n = 31) five minutes to solve a sam- ple GRE math test with five difficult questions while using the MIT pen. Then, participants were asked how confident they would be (two items; 0 to 100 scale; a = .92) and how much effort they would expend (four items, 0 to 100 scale; a = .95) if they were asked to take a GRE test with 30 math questions for 30 minutes. As we expected, a regression analysis revealed that entity theorists (vs. incremental theo- rists) reported a higher level of self-efficacy (b = .39, t(28) = 2.12, p < .05) and intention to exert more effort (b = .43, t(28) = 2.47, p < .05). More importantly, a bootstrap analy- sis using the INDIRECT SPSS macro (Preacher and Hayes 2008) confirmed a significant mediating pathway from implicit self-theories to intended effort on the next GRE test through self-efficacy (95% CI: 1.03, .03). These findings provide added support for the view that brand use can enhance self-efficacy for entity (but not incremental) theo- rists and that increases in self-efficacy lead to greater effort being expended to perform well in a challenging task. Finally, we contribute to an understanding of individual differences in the way brands influence behavior. Prior research on placebo effects and nonconscious brand priming (two major lines of research providing evidence that brands can affect behavior) has shown that brands influence some consumers more than others. For example, brand priming research has reported that incidental exposure to brand logos unconsciously elicits goal-directed behavior, but the effect is stronger for consumers with a stronger motivation to achieve the goals that brands activate (Fitzsimons, Char- trand, and Fitzsimons 2008). In addition, placebo studies reveal that brands affect behavior on a nonconscious level, but the effect is most evident for consumers who are moti- vated to experience the benefits of using the branded prod- uct (Irmak, Block, and Fitzsimons 2005). Our research also indicates individual differences, showing that brand use affects subsequent behavior for entity, but not incremental, theorists. Entity theorists rely on brands as a source of self- efficacy, which can be interpreted as being more motivated to depend on and experience the benefits the brand promise offers, as opposed to incremental theorists, who are more motivated to develop their own skills and abilities as a way to develop self-efficacy. Thus, a consistent theme across all three lines of research is that individual differences exist in the degree to which brands influence behavior. Managerial Implications Our findings suggest that different promotional strategies may be effective for entity versus incremental theorists. Most firms promote the functional, emotional, or self- expressive aspects of their brands. For entity theorists, these brand elements should be cast as assuring consumers that they can achieve, perform better at, or succeed at challeng- ing tasks relevant to the brand. Advertising copy, consumer testimonials, or even brand slogans could contain these sen- timents. Examples of this execution strategy can be found in slogans such as You can do it. We can help (Home Depot) and Be all that you can be (U.S. Army). These slogans should appeal to entity theorists, although entity theorists would need to actually use the brand to receive the prom- ised benefits (as Study 2 shows). In contrast, for incremen- tal theorists, the emphasis should be on providing informa- tion that affords a learning opportunity. Examples of this strategy would be tips and videos for getting the best per- formance from products. Although these types of tips are often included with products, our findings suggest that incremental theorists (who believe they can develop their abilities) are most likely to find these tips appealing. Firms could target these different promotional strategies to entity and incremental theorists in several ways. First, a customer database or online customer profile could include measures of implicit self-theories. Then, advertising appeals could be tailored for entity versus incremental theorists and delivered through e-mail, website links, or pop-up windows. Another option would be to prime consumers to be entity theorists through ad copy and images consistent with entity theory beliefs, in line with our manipulation of implicit self- theories in Study 2. Then, the firm could deliver communi- cations appealing to an entity or incremental theory orienta- tion to the entire customer base. Future Research Directions Our findings suggest several directions for further research. First, our article focuses on general implicit self- theories that are applicable to overall personality domains (Levy, Stroessner, and Dweck 1998; Plaks, Grant, and Dweck 2005). Further research could examine whether con- sumer brand use is also affected by holding domain-specific implicit self-theories such as intelligence (Erdley et al. 1997; Robins and Pals 2002), morality (Chiu, Hong, and Dweck 1997; Dweck and Leggett 1988), shyness (Beer 2002), self-control (Mukhopadhyay and Johar 2005), and relationships (Knee 1998). For example, research has found that self-control lay theories, which refer to the amount of self-control people believe they have (unlimited vs. limited) and whether it can be changed over time (malleable vs. fixed), moderate the effect of self-efficacy on performance (Mukhopadhyay and Johar 2005). When people hold the belief that self-control is limited but malleable, self-efficacy beliefs predict performance well. Extending this line of research to our context, it would be worthwhile to examine whether self-control lay theories moderate the effect of brand use on task performance for entity theorists. Second, we might ask whether entity theorists become less reliant on brand use to heighten their sense of self-efficacy as they accumulate experiences with challenging tasks. Prior research has shown that entity theorists exhibit decreasing patterns of self-efficacy as they are exposed to difficult situations repeatedly (Robins and Pals 2002). We speculate that if entity theorists are able to use brands as a crutch to enhance their sense of self-efficacy, this inter- vention will interrupt a recursive process of worsening self- efficacy and performance and may result in a more stable set of self-efficacy beliefs and performance over time. Fur- thermore, as entity theorists develop self-efficacy beliefs over time, they may adopt more incremental theory beliefs. Thus, entity theorists may become less reliant on brand use. Third, researchers might examine other variables that increase the self-efficacy effect. In our studies, the branded items we used were designed to strip away functional bene- fits associated with the brand. For example, recall that the Gatorade water was tap water poured into a Gatorade logo cup. What would happen if participants used a branded item with which they could also experience the brands func- tional aspects that can improve their abilities? We would expect that self-efficacy effects might be the same or a bit stronger for entity theorists. However, the real difference could emerge for incremental theorists, who might experi- ence feelings of self-efficacy from observing actual improvement in their abilities from the functional aspects of the brand. An intriguing possibility is that all consumers derive a sense of self-efficacy from using brands, but entity theorists respond to the assurance of better performance they receive from brand use, whereas incremental theorists respond to functional aspects of the brand. The pursuit of these lines of inquiry holds the promise of delivering insights into the way consumers interact with brands and how this interaction influences their own behav- ior. Studies that incorporate actual brand use, similar to our research, can be an important addition to prior work that has typically stopped short of brand use. Including actual prod- uct use in more consumer research will provide a new con- ceptual frame for understanding how consumer beliefs and behaviors shape, and are shaped by, their consumption experiences. REFERENCES Aiken, Leona S. and Stephen G. West (1991), Multiple Regression: Testing and Interpreting Interactions. Thousand Oaks, CA: Sage Publications. Bandura, Albert (1982), Self-Efficacy Mechanism in Human Agency, American Psychologist, 37 (2), 12247. (1986), Social Foundations of Thought and Action: A Social Cognitive Theory. Englewood Cliffs, NJ: Prentice Hall. (1988), Reflection on Nonability Determinants of Compe- tence, in Competence Considered: Perceptions of Competence and Incompetence Across the Lifespan, Robert J. Sternberg and John Kolligian Jr., eds. New Haven, CT: Yale University Press, 31562. (1993), Perceived Self-Efficacy in Cognitive Development and Functioning, Educational Psychologist, 28 (2), 11748. (1995), Self-Efficacy: The Exercise of Control. New York: Freeman. (2006), Guide for Constructing Self-Efficacy Scales, in Self-Efficacy Beliefs of Adolescents, Vol. 5, Frank Pajares and Tim Urdan, eds. Greenwich, CT: Information Age Publishing, 307337. and Daniel Cervone (1983), Self-Evaluative and Self- Efficacy Mechanisms Governing the Motivational Effects of Goal Systems, Journal of Personality and Social Psychology, 45 (5), 101728. and Robert E. Wood (1989), Effect of Perceived Controlla- bility and Performance Standards on Self-Regulation of Com- plex Decision Making, Journal of Personality and Social Psy- chology, 56 (5), 805814. Beer, Jennifer S. (2002), Implicit Self-Theories of Shyness, Journal of Personality and Social Psychology, 83 (4), 1009 1024. Chartrand, Tanya L., Joel Huber, Baba Shiv, and Robin J. Tanner (2008), Nonconscious Goals and Consumer Choice, Journal of Consumer Research, 35 (2), 189201. Chiu, Chi-Yue, Ying-Yi Hong, and Carol S. Dweck (1997), Lay Dispositionism and Implicit Theories of Personality, Journal of Personality and Social Psychology, 73 (1), 1930. Cohen, Jacob and Patricia Cohen (1983), Applied Multiple Regres- sion/ Correlation Analyses for the Behavioral Sciences. Hills- dale, NJ: Lawrence Erlbaum Associates. Dweck, Carol S. (2000), Self-Theories: Their Role in Motivation, Personality and Development. Philadelphia: Psychology Press. and Ellen L. Leggett (1988), A Social-Cognitive Approach to Motivation and Personality, Psychological Review, 95 (2), 25673. Elliott, Elaine S. and Carol S. Dweck (1988), Goals: An Approach to Motivation and Achievement, Journal of Person- ality and Social Psychology, 54 (1), 512. Erdley, Cynthia A., Kathleen M. Cain, Catherine C. Loomis, Frances Dumas-Hines, and Carol S. Dweck (1997), Relations Among Childrens Social Goals, Implicit Personality Theories, and Responses to Social Failure, Developmental Psychology, 33 (2), 26372. Feldman, Jack M. and John G. Lynch Jr. (1988), Self-Generated Validity and Other Effects of Measurement on Belief, Attitude, Intention, and Behavior, Journal of Applied Psychology, 73 (3), 42135. Fitzsimons, Grainne M., Tanya L. Chartrand, and Gavan J. Fitzsi- mons (2008), Automatic Effects of Brand Exposure on Moti- vated Behavior: How Apple Makes You Think Different, Jour- nal of Consumer Research, 35 (1), 2135. 246 JOURNAL OF MARKETING RESEARCH, APRIL 2014 Brand Use, Self-Efficacy, and Performance 247 Gist, Marilyn E. and Terence R. Mitchell (1992), Self-Efficacy: A Theoretical Analysis of Its Determinants and Malleability, Academy of Management Review, 17 (2), 183211. Irmak, Caglar, Lauren G. Block, and Gavan J. Fitzsimons (2005), The Placebo Effect in Marketing: Sometimes You Just Have to Want It to Work, Journal of Marketing Research, 42 (Novem- ber), 406409. Jones, Edward E., Frederick Rhodewalt, Steven Berglas, and James A. Skelton (1981), Effects of Strategic Self-Presentation on Subsequent Self-Esteem, Journal of Personality and Social Psychology, 41 (3), 407421. Kramer, Thomas and Lauren Block (2008), Conscious and Non- Conscious Components of Superstitious Beliefs in Judgment and Decision-Making, Journal of Consumer Research, 34 (6), 78393. Levy, Sheri R., Steven J. Stroessner, and Carol S. Dweck (1998), Stereotype Formation and Endorsement: The Role of Implicit Theories, Journal of Personality and Social Psychology, 74 (6), 142136. Lickel, Brian, David L. Hamilton, and Steven J. Sherman (2001), Elements of a Lay Theory of Groups: Types of Groups, Rela- tional Styles, and the Perception of Group Entitativity, Person- ality and Social Psychology Review, 5 (2), 12940. Martocchio, Joseph J. (1994), Effects of Conceptions of Ability on Anxiety, Self-Efficacy, and Learning in Training, Journal of Applied Psychology, 79 (6), 81925. Mukhopadhyay, Anirban and Gita Venkataramani Johar (2005), Where There Is a Will, Is There a Way? The Effects of Con- sumers Lay Theories of Self-Control on Setting and Keeping Resolutions, Journal of Consumer Research, 31 (4), 77986. Muraven, Mark, Dianne M. Tice, and Roy F. Baumeister (1998), Self-Control as Limited Resource: Regulatory Depletion Pat- terns, Journal of Personality and Social Psychology, 74 (3), 77489. Pajares, Frank and Dale H. Schunk (2001), Self-Beliefs and School Success: Self-Efficacy, Self-Concept, and School Achievement, in International Perspectives on Individual Dif- ferences: Self-Perception, Vol. 2, Richard J. Riding and Stephen Rayner, eds. Westport, CT: Ablex, 23965. Park, Ji Kyung and Deborah Roedder John (2010), Got to Get You into My Life: Do Brand Personalities Rub Off on Con- sumers? Journal of Consumer Research, 37 (4), 65569. Plaks, Jason E., Heidi Grant, and Carol S. Dweck (2005), Viola- tions of Implicit Theories and the Sense of Prediction and Con- trol: Implications for Motivated Person Perception, Journal of Personality and Social Psychology, 88 (2), 24562. Preacher, Kristopher and Andrew F. Hayes (2008), Asymptotic and Resampling Strategies for Assessing and Comparing Indi- rect Effects in Multiple Mediator Models, Behavior Research Methods, 40 (3), 87991. Rhodewalt, Frederick (1994), Conceptions of Ability, Achieve- ment Goals, and Individual Differences in Self-Handicapping Behavior: On the Application of Implicit Theories, Journal of Personality, 62 (1), 6785. Robins, Richard W. and Jennifer L. Pals (2002), Implicit Self- Theories in the Academic Domain: Implications for Goal Orien- tation, Attributions, Affect, and Self-Esteem Change, Self and Identity, 1, 31336. Schneider, John A. and W. Stewart Agras (1985), A Cognitive Behavioural Group Treatment of Bulimia, British Journal of Psychiatry, 146 (1), 6669. Shiv, Baba, Ziv Carmon, and Dan Ariely (2005), Placebo Effects of Marketing Actions: Consumers May Get What They Pay For, Journal of Marketing Research, 42 (November), 38393. Singer, Judith D. (1998), Using SAS Proc Mixed to Fit Multilevel Models, Hierarchical Models, and Individual Growth Models, Journal of Educational and Behavioral Statistics, 24 (4), 323 55. Watson, David, Lee A. Clark, and Auke Tellegen (1988), Devel- opment and Validation of Brief Measures of Positive and Nega- tive Affect: The PANAS Scales, Journal of Personality and Social Psychology, 54 (6), 106370. West, Stephen G., Leona S. Aiken, and Jennifer L. Krull (1996), Experimental Personality Designs: Analyzing Categorical by Continuous Variable Interactions, Journal of Personality, 64 (1), 148. Wood, Robert E. and Albert Bandura (1989), Social Cognitive Theory of Organizational Management, Academy of Manage- ment Review, 14 (3), 36184. Yorkston, Eric A., Joseph C. Nunes, and Sashi Matta (2010), The Malleable Brand: The Role of Implicit Theories in Evaluating Brand Extensions, Journal of Marketing, 74 (January), 8093. Zimmerman, Barry J., Albert Bandura, and Manuel Martinez-Pons (1992), Self-Motivation for Academic Attainment: The Role of Self-Efficacy Beliefs and Personal Goal Setting, American Educational Research, 29 (3), 66376.