Performance Brand Placebos: How Brands Improve Performance and Consumers Take the Credit - Digital Wellbeing
←
→
Page content transcription
If your browser does not render page correctly, please read the page content below
Performance Brand Placebos: How Brands
Improve Performance and Consumers
Downloaded from https://academic.oup.com/jcr/article-abstract/42/6/931/2358047 by UNIVERSITY OF THE ARTS LONDON user on 28 February 2019
Take the Credit
AARON M. GARVEY
FRANK GERMANN
LISA E. BOLTON
This research examines how consumption of a performance branded product sys-
tematically improves objective outcomes in a variety of contexts. Five field and labo-
ratory studies demonstrate that this performance brand effect emerges through psy-
chological mechanisms unrelated to functional product differences, consistent with a
placebo. Furthermore, whereas this effect emerges only when there is an expecta-
tion that the performance branded product affects outcomes, consumers attribute
gains to themselves. The performance brand placebo is due to a lowering of task-
induced anxiety, driven by heightened state self-esteem. Several theoretically rele-
vant boundaries are revealed. Stress mindset moderates the effect, strengthening
with the belief that stress is debilitating and weakening (to the point of reversal) with
the belief that stress is enhancing. Moreover, those consumers lower in preexisting
domain self-efficacy beliefs exhibit more substantial performance gains, whereas for
those particularly high in domain self-efficacy, the placebo is mitigated.
Keywords: brands, placebo, performance anxiety, stress mindset
F irms frequently promise consumers that use of their
brands will improve performance outcomes. From the
middle-school child considering the premier brands of soc-
from such brands remains “you will perform better with us.”
Firms often construct compelling arguments as to why their
performance brands are effective at improving performance.
cer shoes, to the college graduate weighing which graduate Claims of superior materials, craftsmanship, design, or other
test prep course to take, a ubiquitous marketing message components can be quite convincing, and certainly true in
some instances. However, when products or services are
Aaron M. Garvey is assistant professor of marketing at the Gatton functionally homogeneous, could the simple belief that a
College of Business and Economics, University of Kentucky, Lexington, particular brand is effective at enhancing performance actu-
KY 40506, (859) 257-2869, aarongarvey@uky.edu; Frank Germann is as- ally improve performance objectively? To answer this ques-
sistant professor of marketing at the Mendoza College of Business, tion, we develop and empirically validate a framework for
University of Notre Dame, Notre Dame, IN 46556, (574) 631-4858, fger-
mann@nd.edu; Lisa E. Bolton is professor of marketing at the Smeal
performance brand consumption by drawing on the litera-
College of Business, Pennsylvania State University, University Park, PA ture examining placebo effects (Plassmann et al. 2008; Shiv,
16802, (814) 865-4175, boltonle@psu.edu. The first two authors contrib- Carmon, and Ariely 2005; Waber et al. 2008), performance
uted equally to this research. The authors would like to thank the editor, anxiety and stress responses (Crum, Salovey, and Achor
associate editor, and four anonymous reviewers for their helpful com- 2013; Eysenck et al. 2007), and self-attributions (Malle
ments and suggestions. The authors also thank participants in seminars at 2006). In doing so, our research increases knowledge at the
the University of Kentucky, University of Notre Dame, Pennsylvania intersection between branded consumption and consumer
State University, University of Cincinnati, and the WHU - Otto Beisheim performance outcomes. Doing so also helps address whether
School of Management for their helpful feedback on this research. premium sneakers or test prep courses for the student
Darren Dahl served as editor, and Page Moreau served as associate edi- athletes of your family represent wise investments.
tor for this article. Performance brand offerings—branded goods and
services expected to enhance personal performance
Advance Access publication December 28, 2015
outcomes—span a variety of multibillion dollar industries.
C The Author 2015. Published by Oxford University Press on behalf of Journal of Consumer Research, Inc.
V
All rights reserved. For permissions, please e-mail: journals.permissions@oup.com Vol. 42 2016
DOI: 10.1093/jcr/ucv094
931932 JOURNAL OF CONSUMER RESEARCH
In the United States alone, exam preparation products and stress and anxiety may either enhance or debilitate depend-
services represent a $7.3 billion industry, athletic apparel ing on consumer belief (Brooks 2013; Crum et al. 2013).
$9.9 billion, and overall sporting goods a staggering $63 As a second moderator, we examine the impact of domain
billion (Barnes Reports 2013; Statistica Dossier 2014). self-efficacy beliefs (i.e., preexisting personally held be-
Downloaded from https://academic.oup.com/jcr/article-abstract/42/6/931/2358047 by UNIVERSITY OF THE ARTS LONDON user on 28 February 2019
Surprisingly, how the consumption of performance brands liefs about one’s capability to produce attainments in a spe-
within these industries influences objective performance is cific domain; Bandura 1997, 2006) on the strength of the
not well understood theoretically and has received rela- performance brand placebo. Consistent with an anxiety-re-
tively little attention in the consumer behavior literature. duction account, individuals holding unfavorable views of
Our research focuses on the implications of performance their own domain self-efficacy (who experience heightened
brand consumption for the consumer’s mental and emo- task anxiety; Bandura 1991) receive a greater objective
tional state in influencing task outcomes, rather than the boost from the performance brand placebo effect. As a
material differences that such brands may provide. We third moderator, we refine the performance brand construct
demonstrate that performance brand consumption has ob- by distinguishing performance brands from prestige brands
jectively measurable effects on performance despite illu- (Park, Milberg, and Lawson 1991; Wilcox, Kim, and Sen
sory (i.e., immaterial) brand differences, consistent with a 2009) that may be highly regarded by consumers but do
placebo effect (Shiv et al. 2005). In doing so, our research not carry strong associations of positive performance ex-
contributes to the literature in five ways. pectations and therefore do not drive a performance
First, our work expands research examining positive pla- placebo.
cebo effects beyond subjective outcomes (e.g., perceived Fourth, this work expands understanding of brand-
pain reduction) to explain how actual objective outcomes related consumer attribution processes. Our work reveals
are systematically improved or harmed by performance that consumers do not give performance brands credit for
brand consumption. Whereas recent research has docu- the performance boost. Although use of performance
mented a placebo that undermines performance due to mar- brands can lead to better outcomes for consumers, both in
keting actions (e.g., Shiv et al. 2005; Wright et al. 2013), terms of personal performance and self-esteem, we find
little consumer research exists on placebos that enhance that consumers credit the boost to themselves rather than
objective performance. This lack of understanding exists de- the brand. Previous research has argued that consumers
spite the multibillion dollar global industries around brand- purchase brands to help construct their self-concept and
driven performance products. By exploring the intersection bolster self-esteem (Reimann and Aron 2009), which can
of brand consumption and consumer performance, our work improve consumer-brand connections (Escalas and
begins to address these theoretical and substantive gaps. Bettman 2005). However, our research demonstrates that
Second, we shed insight into the psychological under- performance outcomes are attributed to both the brand and
pinnings of the performance brand placebo by proposing the self, with credit for the boost in performance outcomes
and providing empirical support for an anxiety-reduction going to consumers themselves. In that regard, consumers
mechanism. Specifically, a performance brand is shown to can be said to give only partial rather than full credit to the
improve state self-esteem and, in turn, reduce stress-in- brand for performance outcomes.
duced anxiety, thereby enhancing performance. In doing Finally, our findings have implications for marketers
so, we expand understanding of the implications of and consumers of performance brands. Our findings that
branded consumption for related stress responses. performance brands enhance consumer proficiency but do
Furthermore, we demonstrate that the psychological under- not receive credit is ironic in light of recent research and
pinnings of brand-driven enhancing performance placebos criticism suggesting that brand premiums in functionally
are fundamentally different in nature from those of tradi- homogeneous product categories lead to wasteful spending
tional subjective placebo effects documented in the litera- that harms consumers (Bronnenberg et al. 2014). Indeed,
ture (Hr objartsson and Gøtzsche 2004). our research suggests that such performance brands, absent
Third, we identify theoretically and pragmatically rele- functional differences, may provide objective benefits that
vant moderators that provide boundaries for the placebo ef- help consumers. Consumers’ failure to fully acknowledge
fect on objective performance. As a first moderator, the performance brand’s contribution creates a dilemma
consistent with an anxiety-reduction mechanism for the for marketers wishing to receive more of the credit for ful-
placebo, we demonstrate the moderating role of individual filling their promise of better performance. These and other
stress mindset (i.e., whether stress has a debilitating or en- managerial implications are expanded on in the General
hancing effect on individuals; Crum et al. 2013). Discussion section.
Specifically, the positive performance placebo strengthens
with the belief that stress is debilitating and weakens (to THEORETICAL FRAMEWORK
the point of reversal) with the belief that stress is enhanc-
ing. Doing so also contributes to the emerging literature The present research focuses on the impact of consum-
examining how the emotional arousal associated with ing performance brands on individuals in accomplishingGARVEY ET AL 933
outcomes, in the absence of material product differences. with a classic study by Branthwaite and Cooper (1981) in
Admittedly, there likely do exist material, substantial dif- which women who received a sugar pill positioned as a
ferences between many performance branded products premium aspirin brand reported greater headache relief
available to consumers. For example, the specialized con- versus women consuming the same sugar pill positioned as
Downloaded from https://academic.oup.com/jcr/article-abstract/42/6/931/2358047 by UNIVERSITY OF THE ARTS LONDON user on 28 February 2019
struction process to create a Louisville Slugger may result unbranded. Similarly, Plassmann et al. (2008) demon-
in greater balance and rebound against a baseball, thereby strated that the pleasure experienced from consuming a
improving batting performance. However, such actual wine was greater when consumers believed that wine to be
functional differences associated with performance brands higher priced.
are not the focus of our present research. Rather, we focus The finding that subjective interpretations of product ex-
on the impact of performance brands in changing profi- perience can assimilate to expectations has been long es-
ciency at the target endeavor, absent any material or func- tablished in behavioral literature (Herr, Sherman, and
tional differences. Such an improvement in performance Fazio 1983; Hoch and Ha 1986) and provides an intuitive
due to illusory (i.e., immaterial) brand differences is con- and reasonable mechanism for the emergence of subjective
sistent with a placebo effect. placebos. Indeed, given that premium prices and brands
typically carry with them heightened expectations, the
emergence of positive placebos for subjective outcomes
Performance Brand Placebos seems to directly follow. Whether a similar effect emerges
What is a placebo? A formal definition has been the topic in unambiguous, objective performance outcomes is less
of substantial debate (Moerman and Jonas 2002). certain, however. Put simply, can marketing assets and ac-
Chaucer’s character, Placebo, is a shameless flatterer who tions such as brands and price changes lead to actual per-
bolsters and strengthens the confidence of the vain formance changes, that is, cause performance placebo
Januarie in The Canterbury Tales. Claims about placebos effects?
have been frequent and far-ranging in the medical literature
over time, dating back in modern form nearly a century Objective Performance. Important evidence for an ob-
(Kerr, Milne, and Kaptchuk 2008). Within the medical lit- jective performance placebo was first established by Shiv
erature, the placebo has been defined as “a substance or et al. (2005), who demonstrated that discounting the price
procedure that is without specific activity for the condition of an energy drink resulted in lower performance on puzzle
being treated,” and the placebo effect is any therapeutic ef- tasks versus a full-priced alternative. This work established
fect produced by such a placebo (Shapiro and Shapiro a performance-diminishing placebo (driven by price dis-
1997). Furthermore, this inert substance is presented to the counts), and it identified product expectancies as important
recipient as an active substance, that is, a “sham proce- to the emergence of the performance placebo. Wright et al.
dure” (Finniss et al. 2010). Thus a placebo effect is any (2013) replicated Shiv et al.’s (2005) price discount-driven
measurable difference between a control group and a sepa- performance placebo effect, and Amar et al. (2011) ob-
rate group that receives the exact same treatment (product/ served a relationship between brand reputation and product
service) but believes the treatment to be fundamentally effectiveness. Neither work expanded on underlying pro-
different. cesses for performance diminishment, whereas Irmak,
Block, and Fitzsimons (2005) point to the role of motiva-
Subjective Placebo Outcomes. A meta-analysis of medi- tion in placebo effects. As Shiv et al. (2005) acknowledge,
cal placebo effects since 1946 conducted by Hrobjartsson and “Given the substantial power and robustness of placebo ef-
Gøtzsche (2004) found that placebos did significantly im- fects, these effects are most likely multiply determined.”
prove subjective outcomes (e.g., self-reports of experienced Against this backdrop, the present research posits that
pain) but had no significant effects on objectively measurable brands can elicit performance placebo effects, and, further-
outcomes (e.g., hypertension). That is, placebos did not result more, that objective performance may actually be en-
in distinguishable physiological outcomes versus a no-pla- hanced through consumption of a performance brand.
cebo control. Thus if brand-driven performance placebos ex- Whereas prior research has demonstrated that marketing
ist, it is unlikely that these emerge due to direct physical actions that reduce product expectancies may undermine
changes to the consumer, but rather because use of the brand objective performance, we propose that brands can carry
alters some aspect of the consumer’s mental state at the time positive expectancies that improve performance outcomes.
of performance. In this research, we define “performance brands” as
Consumer behavior research examining placebo effects branded goods and services that carry strong, positive per-
due to marketing actions have also predominantly observed formance expectancies specific to a task or set of tasks. We
subjective placebo effects. For example, Waber et al. also explore the underlying mechanisms for a perfor-
(2008) demonstrated that individuals who consumed a dis- mance-enhancing brand placebo, proposing that the mental
counted analgesic reported experiencing more pain from state of the placebo recipient plays a critical role.
electrical shocks to the wrist. These results are consistent Specifically, we theorize that the performance brand934 JOURNAL OF CONSUMER RESEARCH
placebo emerges because consumption of a performance 1974), and consumption of a performance brand should ac-
brand reduces experienced stress and associated maladap- tivate a schema that is either congruent with a preexisting
tive anxiety by bolstering state self-esteem. positive self-image (of proficiency) or congruent with an
idealized self (Belk 1988; Berger and Ward 2010; Fournier
Downloaded from https://academic.oup.com/jcr/article-abstract/42/6/931/2358047 by UNIVERSITY OF THE ARTS LONDON user on 28 February 2019
Anxiety Reduction and Boosted State 1998; Muniz and O’Guinn 2001; Sirgy 1982). To the ex-
tent that the brand is congruent with the perceived or ideal-
Self-Esteem ized self, it should reinforce that positive self-view and
State anxiety typically arises from a stressful situation in accordingly bolster self-esteem (Ferarro et al. 2011).
which outcomes are uncertain or uncontrollable, when per- Moreover, heightened self-esteem is closely related to
formance evaluation is a certainty, or when the individual anxiety, such that anxiety decreases as self-esteem rises
otherwise perceives a threat to the self (Derakshan and (whether trait or state) (Baumeister et al. 2003; Brockner
Eysenck 2009; Raghnunathan and Pham 1999). More gen- 1983; Heatherton and Polivy 1991; Leary et al. 1995;
erally, state anxiety has been described as an outcome of Pyszczynski et al. 1989; Tennen and Herzberger 1987).
experienced stress (Duhacheck 2005; Friedman, Clark, and For example, Greenberg et al. (1992) manipulated state
Gershon 1992). The experience of state anxiety has been self-esteem through feedback on a personality test: partici-
demonstrated to redirect attention and cognitive resources pants then exposed to stressors (e.g., an artificially low IQ
away from consciously pursued outcomes (Eysenck et al. test result or alarming image) reported less anxiety when
2007) and to increase ruminative thought (Carver and their self-esteem had been heightened. Furthermore,
Scheier 1988), thereby interfering with performance in heightened self-esteem can serve to attenuate task-induced
achieving those outcomes. As such, the negative impact of anxiety, both in competitive and noncompetitive contexts.
anxiety on physical and cognitive performance is well doc- For example, elite Swedish athletes training for the
umented across a variety of disciplines including verbal Olympics demonstrated lower anxiety during competitive
and mathematic test taking (Ashcraft and Faust 1994; performances as their self-esteem increased (Koivula,
Elliot and McGregor 1999), athletic competition (Hall and Hassmén, and Fallby 2002). Similarly, state anxiety among
Kerr, 1998; Hanton, Mellalieu, and Hall 2002), performing professional orchestral and student musicians has been
arts such as music (Deen 2000; Ryan 2004), dance demonstrated to decrease as self-esteem increases
(Tamborrino 2001), and acting (Wilson 2002), and even (Langendörfer et al. 2006; Sinden 1999).
sexual performance (McCabe 2005) and public speaking In summary, we predict that consumption of a perfor-
(Brooks 2013; Merritt, Richards, and Davis 2001) (for a re- mance brand lowers the consumer’s task-induced anxiety
view also Eysenck 1992; Spencer, Steele, and Quinn 1999; and thus improves his or her objective performance due to
Steele 1997; Steele and Aronson 1995; Stone et al. 1999). heightened state self-esteem. Formally:
For example, within the realm of athletic competition,
H1: Consumption of a performance brand improves objec-
Burton (1988) observed that swimmers higher in anxiety
tive performance outcomes.
immediately prior to a competition race swam slower ver-
sus expectations. H2: The performance brand placebo effect proposed in hy-
Given this relationship, strategies that reduce anxiety pothesis 1 is mediated by (a) heightened state self-esteem
should therefore improve objective performance—as re- that (b) subsequently lowers anxiety.
search in a variety of contexts attests (cf. Eysenck et al.
2007). For example, Algaze (1995) demonstrated that a
workshop intervention aimed at reducing academic anxiety
resulted in improved performance. However, research has
Attributions for Performance Gains
not to our knowledge examined the implications of Hypotheses 1 and 2 predict that consumption of a perfor-
branded product consumption in reducing anxiety and mance brand enhances objective performance, which raises
thereby enhancing performance—the focus of the present an interesting question: To what extent do consumers at-
research. Specifically, we theorize that task-related anxiety tribute performance to the brand versus the self? On the
may be reduced by the consumption of a performance one hand, use of a performance brand could draw attention
brand. We also develop theory that suggests this reduction to the brand. Inasmuch as attributions are inaccurate and
is due to a bolstering of the consumer’s state self-esteem tend to be driven by salient factors (cf. Kelley 1973), im-
(consistent with an affectively felt sense of personal worth; proved performance may be attributed to the performance
Ferarro, Escalas, and Bettman 2011; Heatherton and Polivy brand itself. Similarly, branded consumption has been
1991; Pelham and Swann 1989), which stems from the linked to heightened connection with that brand (Escalas
consumption of a performance brand. and Bettman 2005), which also suggests that the connected
Brand use is closely linked to consumer perceptions of brand will receive credit for performance enhancement. On
the self (Berger and Heath 2007; Dolich 1969; Escalas and the other hand, research has argued that placebo effects in
Bettman 2005; Kleine, Kleine, and Kernan 1993; Landon general occur largely outside of conscious awareness (ShivGARVEY ET AL 935
et al. 2005; Stewart-Williams and Podd 2004), potentially placebo benefits to consumers that include objective per-
decreasing the likelihood of attributions to the brand. In ad- formance outcomes, thereby living up to its brand
dition, our work argues that the performance brand will promise—but consumers will downplay the performance
heighten state self-esteem, which could lead consumers to brand’s contribution and credit themselves with heightened
Downloaded from https://academic.oup.com/jcr/article-abstract/42/6/931/2358047 by UNIVERSITY OF THE ARTS LONDON user on 28 February 2019
infer that the self is primarily responsible for positive out- performance.
comes. For example, heightened self-esteem has been
linked positively to self-serving biases (e.g., Blaine and
Crocker 1993), self-delusions (e.g., Colvin, Block, and EMPIRICAL OVERVIEW
Funder 1995), and narcissism (e.g., Jordan et al. 2003).
Moreover, attributions for positive outcomes tend to be re- A series of studies was conducted to test our hypothe-
markably self-serving (Bradley 1978; Malle 2006; Miller ses; Figure 1 provides an organizing framework. Study 1
and Ross 1975). Indeed, Fitch (1970) showed that subjects examines the impact of performance brands in an athletic
attribute significantly more causality to internal sources for context and provides preliminary evidence for a positive
success outcomes than for failure outcomes; importantly, placebo on objective performance outcomes. Study 2 ex-
this effect strengthened as self-esteem was enhanced. plores the underlying psychological process (specifically,
Given that the performance brand enhances state self-es- the role of state self-esteem), as well as the downstream
teem, we predict that consumers will become more likely consequences for consumer attributions regarding perfor-
to attribute performance to themselves. In contrast, we ex- mance. Study 3 details the process through which anxiety
pect that acknowledgment of the brand’s contribution will reduction due to heightened state self-esteem improves
not similarly increase. Unlike the prediction for perfor- performance outcomes, and it provides evidence that the
mance outcomes in hypothesis 2, we expect mediation of performance brand placebo (and corresponding attribu-
performance brand effects via state self-esteem (but not tions to the self) generalizes to cognitive tasks. Studies 4
anxiety) inasmuch as enhanced self-esteem can improve at- and 5 further refine our theory by exploring boundary
tributions to the self directly. Formally: conditions for the performance brand placebo effect; hy-
potheses are introduced with each study. Specifically,
H3: Consumption of a performance brand (a) increases per- study 4 further demonstrates the underlying role of anxi-
formance attributions in favor of the self by (b) heightening ety via stress mindset as a moderator. Finally, study 5
state self-esteem. demonstrates (1) the moderating role of domain self-effi-
cacy beliefs and also (2) distinguishes performance
If supported, hypotheses 1, 2 and 3 pose an interesting brands from other brands that are highly regarded by con-
paradox for marketers: A performance brand may provide sumers (i.e., prestige brands). Together, the set of findings
FIGURE 1
ORGANIZING FRAMEWORK
Moderators:
Stress Mindset (Study 4; H4)
Prestige (vs. Performance) Brands (Study 5; H5)
Self-Efficacy Beliefs (Study 5; H6)
Heightened
Performance Brand Lowered Improved
State Self-
Consumption State Anxiety Performance
(Studies 1-5) Esteem (Studies 3,4) (Studies 1-5)
(Studies 2,3)
Increased Self
Attributions
(Studies 2,3; H3)
NOTE.—The shaded path refers to the positive performance brand placebo (i.e., hypotheses 1 and 2).936 JOURNAL OF CONSUMER RESEARCH
support a performance brand placebo effect on objective putting task, participants responded to background ques-
outcomes that is consistent with our theoretical account tions (e.g., gender, age).
and the important role played by consumer beliefs (about
brands, the self, and anxiety) when using performance
Results
Downloaded from https://academic.oup.com/jcr/article-abstract/42/6/931/2358047 by UNIVERSITY OF THE ARTS LONDON user on 28 February 2019
brands.
Pretest. A pretest of performance expectancies was
conducted to assess our operationalizations of the strong
STUDY 1: A POSITIVE BRAND PLACEBO performance brand (Nike), weak performance brand
EFFECT ON PERFORMANCE (Starter), and control with no brand information. The pre-
test asked 84 students on a large midwestern university
The objective of the first study is to provide evidence campus to rate how a golf putter was expected to influence
for an objective positive performance brand placebo. golf putting performance. All subjects viewed a photo-
The study utilizes an athletic context and examines how graph of the same golf putter with a manipulated brand la-
the use of a performance branded product alters athletic bel and responded to three items measuring performance
performance in a golf putting task. We expect that ath- expectancies (all on a scale from 1 [Not at all] to 7 [Very
letic performance will be enhanced when a strong per- much] adapted from Shiv et al. 2005): “Using this [brand]
formance brand is used (i.e., testing hypothesis 1). golf putter will harm/help my putting performance”; “I feel
Although prior work has demonstrated that marketing that using this [brand] golf putter will be very bad/very
phenomena that make salient performance deficits can good at improving my putting performance”; “To what ex-
undermine performance (i.e., a negative placebo; Shiv tent could using this [brand] golf putter help your putting
et al. 2005), we hypothesize that a brand carrying strong performance?” As anticipated, expectancies (a ¼ .96) were
performance expectancies will enhance performance. To significantly higher for the strong performance brand
evaluate the direction of the performance brand placebo (Nike) putter than the weak performance brand (Starter)
effect, we test the impact of a strong performance brand putter (Mstrong ¼ 5.02, SD ¼ 1.03 vs. Mweak ¼ 3.99,
versus both a weak performance brand and a nonbranded SD ¼ 1.34; F(1, 55) ¼ 9.96, p < .01) and unbranded control
control. (Mstrong vs. Mcontrol ¼ 4.10, SD ¼ 1.36; F(1, 55) ¼ 8.18,
p < .01). Expectancies did not differ between the weak per-
Method formance brand and control (F(1, 54) ¼ .10, p > .75).
These results support our operationalizations.
Participants and Design. The experiment was a 3
group (strong performance brand/weak performance brand/ Objective Performance. We averaged the number of
control) between-subjects design. A total of 95 students strokes each participant took from the three predefined loca-
(35% male) on a large midwestern university campus par- tions. On average, participants took 2.24 strokes to sink the
ticipated in the study for extra course credit. putt (min ¼ 1, max ¼ 4.67, SD ¼ .86; n ¼ 91). Four partici-
pants distributed across conditions who took an excessive
Procedure. Participants were invited by research assis- number of strokes (i.e., > 3 SD from the initial overall
tants into the lab one at a time and were told they would be mean, n ¼ 95) were omitted from subsequent analyses. (The
participating in a market research study about a new proto- pattern of results does not change if these data are retained.)
type golf putter. Participants were randomly assigned to a Performance (i.e., average number of strokes) was ana-
strong or weak performance brand putter (i.e., Nike vs. lyzed as a function of brand condition and revealed a main
Starter brands, based on a pretest described later), or a con- effect of brand (F(2, 88) ¼ 4.19, p < .05). More germane to
trol group condition in which no brand-related information hypothesis 1, a planned contrast indicated that performance
was provided about the putter. Note that all participants outcomes were enhanced through use of the strong perfor-
used the same putter (ruling out differences in actual putter mance brand versus the control (Mstrong ¼ 1.91, SD ¼ .71
performance); however, the putter’s label was manipulated vs. Mcontrol ¼ 2.49, SD ¼ .89; F(1, 88) ¼ 7.55, p < .01).
to reflect the appropriate condition. That is, as expected, fewer strokes were needed to sink
Participants were asked to complete putts on a putting putts with a strong performance brand versus the control.
green from three predefined locations exactly 2, 3, and 4.5 Performance outcomes were likewise enhanced when con-
feet from the hole. They were instructed to try to get the trasting use of the strong versus weak performance brand
ball into the hole using the least number of putts possible (Mstrong vs. Mweak ¼ 2.36, SD ¼ .90; F(1, 88) ¼ 4.49,
(practice putts were not allowed). Participants were asked p < .05). In contrast, the planned contrast of weak versus
to putt again from the respective initial location if the ball control conditions was not significant (F(1, 88) ¼ .41,
did not go into the hole. A research assistant recorded the p ¼ .53). These results are consistent with hypothesis 1 and
number of strokes needed to sink the ball in the hole from an enhancing effect of brands that carry strong, positive
each location as a measure of actual performance. After the performance expectancies.GARVEY ET AL 937
Discussion needed to sink the ball in the hole from each location as a
measure of actual performance.
Study 1 provides evidence for a positive performance
After the putting task, all participants completed a short
brand placebo on objective outcomes. Golf performance
questionnaire. Participants responded to the following
improved (i.e., taking fewer strokes to sink a putt) when a
Downloaded from https://academic.oup.com/jcr/article-abstract/42/6/931/2358047 by UNIVERSITY OF THE ARTS LONDON user on 28 February 2019
questions to measure performance attributions: “How
brand associated with strong athletic performance expecta-
much did each of the following contribute to how well you
tions was used, compared to a weak brand or no brand in-
performed in the putting task: The performance and quality
formation. These results support our theorizing regarding
of the putter” and “My personal athleticism” (each on 7
the performance brand placebo, with objective improve-
point scales with end points “not at all/very much”).
ments of over 20% from using a strong performance brand.
Participants also responded to a measure of state self-es-
teem (adapted from Robins, Hendin, and Trzesniewski
STUDY 2: STATE SELF-ESTEEM 2001): “Please tell us how you felt while putting,” (1) I felt
MEDIATES PERFORMANCE AND good about myself, and (2) my self-esteem was high (each
ATTRIBUTIONS on 7 point scales with end points “strongly disagree/
strongly agree”). Finally, participants answered back-
Study 2 serves two primary objectives. First, we explore ground questions (e.g., gender, age).
the psychological process that underlies the positive effect
of performance brands on objective outcomes.
Specifically, we test the mediating role of state self-esteem Results
in determining performance outcomes (hypothesis 2). Objective Performance. We again averaged the num-
Second, we not only examine objective brand performance ber of putts from the three predefined locations.
(as in study 1), but also how consumers account for this in- Participants averaged 1.93 strokes to sink the putts
crease in performance via attributions to the self (hypothe- (min ¼ 1; max ¼ 4.33; SD ¼ .76; n ¼ 101). We excluded
sis 3). Consistent with our theorizing, we predict that a five subjects distributed across conditions because of ex-
strong performance brand will enhance state self-esteem, cessive number of putts (i.e., > 3 SD from the initial over-
which in turn increases objective performance and also at- all mean, n ¼ 106; the pattern of results does not change if
tributions to the self for performance. We use an athletic these data are retained.)
context (golf putting) to build on the results of study 1. Performance (i.e., average number of strokes) was ana-
lyzed as a function of brand condition. As expected, fewer
Method strokes were needed with a strong performance brand com-
Participants and Design. The experimental design was pared to the control condition (Mstrong ¼ 1.71, SD ¼ .61;
a 2 group (strong performance brand/control) between-sub- Mcontrol ¼ 2.14, SD ¼ .84; F(1, 99) ¼ 8.83, p < .01), consis-
jects design. A total of 106 students (51% male) on a large tent with hypothesis and a performance brand placebo ef-
midwestern university campus voluntarily participated in fect. As in study 1, performance again improved
the study and received $5 compensation. approximately 20% when using a strong performance
brand (compared to no brand information).
Procedure. Students in a class building on a large uni-
versity campus were intercepted and asked if they would Mediation via State Self-Esteem. As expected, partici-
like to volunteer to participate in a market research study. pants’ state self-esteem (r ¼ .90) was significantly greater
Those who agreed were invited into the lab one at a time. in the strong brand versus the control condition
As in study 1, participants were told that they would be (Mstrong ¼ 5.12, SD ¼ 1.04; Mcontrol ¼ 4.37, SD ¼ 1.19; F(1,
participating in a study about a new prototype golf putter 99) ¼ 11.34, p < .01). This pattern holds if we control for
and were randomly assigned to either the strong perfor- objective performance (F(1, 98) ¼ 4.56; p < .05), helping
mance brand putter (i.e., Nike, based on the pretest de- rule out the possibility that self-esteem was enhanced be-
scribed in study 1) or the control group putter condition in cause consumers were able to observe their own perfor-
which no brand-related information was provided about the mance. (In study 3, we further minimize this possibility by
putter. As in study 1, all participants used the same putter using a context in which performance outcomes are not ob-
(ruling out differences in actual putter performance; to im- vious to participants.) That is, the strong brand (compared
prove our confidence in generalizability of the effect, all to no brand information) enhanced state self-esteem.
participants used a different putter than was used in study To assess the mediating role of state self-esteem, we
1), and the putter’s label was again manipulated to reflect conducted a bootstrap analysis (e.g., Preacher and Hayes
the appropriate condition. Participants were asked to com- 2004, 2008; Zhao, Lynch, and Chen 2010) with brand as
plete putts from three predefined locations on a putting the independent variable, actual performance as the depen-
green. The same procedures were followed as in study 1, dent variable, and state self-esteem as the mediator. State
and a research assistant recorded the number of strokes self-esteem emerged as a significant mediator (Indirect938 JOURNAL OF CONSUMER RESEARCH
effect ¼ 0.23, 95% confidence interval [CI] ¼ .457 current study extends our findings to cognitive perfor-
to .094). These results support hypothesis 2 and media- mance in problem-solving tasks (i.e., hypothesis 1).We
tion of the performance brand placebo via enhanced state also test whether the tendency to attribute performance
self-esteem. gains to the self replicates in this context (i.e., hypothesis
Downloaded from https://academic.oup.com/jcr/article-abstract/42/6/931/2358047 by UNIVERSITY OF THE ARTS LONDON user on 28 February 2019
3). The manipulated product was a pair of noise-blocking
Attributions. How did participants account for their foam earplugs positioned to improve concentration on a
performance? Participants were significantly more likely to cognitive test.
attribute their performance to the self (i.e., personal athleti-
cism) when using the strong performance brand versus the
control (Mstrong ¼ 3.98, SD ¼ 1.30; Mcontrol ¼ 3.26, Method
SD ¼ 1.56; F(1, 99) ¼ 6.34, p ¼ .01). Attributions to the
brand did not differ (Mstrong ¼ 4.47, SD ¼ 1.24; Participants and Design. The design was a 2 group
Mcontrol ¼ 4.08, SD ¼ 1.26; F(1, 99) ¼ 2.47, p > .10). That (strong performance brand/control) between-subjects de-
is, participants attributed their improved performance to sign. A total of 91 undergraduate students (76% male) at a
the self (personal athleticism), consistent with hypothesis large midwestern university voluntarily participated for ex-
3a. tra credit in an introductory business course.
A follow-up bootstrap analysis finds that the impact of
performance brand on attributions to the self is mediated Procedure. Participants entered the behavioral labora-
by state self-esteem (indirect effect ¼ .139, 90% CI, .002– tory and were seated separately in individual cubicles con-
.421), consistent with hypothesis 3b. (We note that this in- taining a computer workstation and an opaque sealed
direct effect pattern holds [indirect effect ¼ .142, 95% CI, plastic container that contained the performance brand
.013–.423] when controlling for actual performance as a product, a pair of foam earplugs. Inside the container was a
covariate). That is, the performance brand heightens state sealed plastic bag containing the earplugs manipulated to
self-esteem, which drives consumers to take the credit for either have a strong performance brand (3M) logo or no
improved performance (rather than increasing perfor- brand information. All participants used the same actual
mance attributions to the brand). These results support hy- model of earplugs. The 3M brand was selected based on
pothesis 3. the results of a pretest described later.
Participants were told that the main part of the study was
Discussion a math test, and that during this test they would wear a pair
of foam earplugs to minimize distractions and improve
Study 2 again demonstrates a positive performance concentration. Participants then received the performance
brand placebo on objective outcomes while shedding light brand manipulation by taking the earplugs from the con-
on the process that underlies the effect. Consumption of a tainer and wearing them (stimulus available in the online
strong performance brand enhances state self-esteem, appendix). Participants then proceeded to the performance
which in turn has a positive impact on performance. task, which consisted of five mathematics problems classi-
Furthermore, we reveal that consumers attribute these fied as moderately difficult by the SAT College Board
performance gains to themselves due to enhanced state Preparation Guide in 2014 (e.g., “Samantha is packing for
self-esteem, whereas the performance brand received no a trip. Of the towels in the closet, 6 are brown. She will
additional credit for performance. randomly pick one of the towels to pack. If the probably is
2/5 that the towel she will pick is brown, how many towels
STUDY 3: PROCESS ROLE OF ANXIETY are in the closet?” 15/18/20/30/36).
Participants next responded to process measures of anxi-
The objective of study 3 is twofold. First, study 3 ex- ety (“I felt anxious”), motivation (“I felt motivated”), and
tends our investigation to include the role of anxiety reduc- enjoyment (“I enjoyed myself”), each on 7 point scales,
tion in driving the performance brand placebo. Whereas with end points “not at all/a lot.” The latter measures were
study 2 establishes that a strong performance brand en- included to examine alternative mediating processes. State
hances state self-esteem and thereby improves perfor- self-esteem (“I felt bad about myself” (reverse scored),
mance, we propose that the performance gains due to “My self-esteem was high”; r ¼ .71) was also recorded.
enhanced state self-esteem emerge from a reduction in Participants also responded to performance attribution
task-related anxiety. Specifically, a strong performance measures (“To what extent would you attribute your per-
brand should enhance state self-esteem and, in turn, reduce formance on this test to: your innate intelligence/ability,
anxiety, thereby positively affecting objective performance your education, and the earplugs”; each on 7 point scales
(i.e., hypothesis 2). with end points “not at all/a lot”). Because individuals
Second, our evidence thus far for a performance brand might be reluctant to claim innate intelligence, we included
placebo has emerged for athletic performance, and the education as an additional self-attribution measure. Finally,GARVEY ET AL 939
participants responded to background questions (e.g., gen- CI, 002–189), whereas no indirect effect was observed for
der, age). motivation or enjoyment (90% CIs contained 0).
Attributions. Analysis indicated that the strong perfor-
Results mance brand increased attributions to the self (r ¼ .75;
Downloaded from https://academic.oup.com/jcr/article-abstract/42/6/931/2358047 by UNIVERSITY OF THE ARTS LONDON user on 28 February 2019
Pretest. A pretest was administered to 57 students on a Mcontrol ¼ 4.46, SD ¼ 1.56 vs. Mstrong ¼ 5.07, SD ¼ 1.32; F(1,
large midwestern university campus to indicate how two 89) ¼ 4.02, p < .05). Attributions to the earplugs were unaf-
types of earplugs (i.e., strong performance brand “3M” or fected by brand condition (Mcontrol ¼ 2.49, SD ¼ 1.59 vs.
no brand information in the control condition) were ex- Mstrong ¼ 2.80, SD ¼ 1.50; F(1, 89) ¼ .89; p > .30).
pected to affect concentration on a math test. All partici- Moreover, a bootstrapping analysis with performance brand
pants viewed the same picture of foam earplugs with condition as the independent variable, state self-esteem as
manipulated brand labels. As in study 1, we used the fol- mediator, and attribution to the self as dependent variable in-
lowing three items to measure brand performance expec- dicated a significant indirect effect (indirect effect ¼ .16, 95%
tancies (all on a scale from 1 [Not at all] to 7 [Very much]; CI, .031–.386). (We note that this indirect effect pattern holds
adapted from Shiv et al. 2005): “Wearing these [brand] when controlling for actual performance as a covariate [indi-
earplugs will harm/help my concentration on a math test”; rect effect ¼ .09; 90% CI, 007–.269]. Also, as expected, an
“I feel that wearing these [brand] earplugs is very bad/very analysis testing serial mediation involving anxiety [i.e.,
good at improving my concentration on a math test”; “To brand—state self-esteem—anxiety—attribution] is not sup-
what extent can wearing these [brand] earplugs improve ported; recall that an enhanced state self-esteem is expected
your concentration on a math test?” As anticipated, perfor- to facilitate attributions to the self directly rather than via anx-
mance expectancies (a ¼ .93) were significantly higher for iety.) These results support hypothesis 3a and 3b and replicate
the strong performance brand earplugs than the unbranded study 2: a strong performance brand (versus no brand infor-
control (Mstrong ¼ 4.83, SD ¼ 1.17 vs. Mcontrol ¼ 3.95, mation) heightens state self-esteem, which in turn increases
SD ¼ .95; F(1, 55) ¼ 9.73, p < .01). performance attributions toward the self—with no similar in-
crease in attributions to the performance branded product.
Objective Performance. Performance was measured
via the number of questions correctly answered (out of 5)
on the cognitive performance task. In support of hypothesis Discussion
1, analysis revealed improved objective performance for Study 3 demonstrates that a performance brand height-
the strong performance brand versus control condition ens state self-esteem and, as a result, (1) reduces anxiety
(Mcontrol ¼ 2.38, SD ¼ 1.13 vs. Mstrong ¼ 2.89, SD ¼ 1.06; and thereby improves performance, and (2) increases con-
F(1, 89) ¼ 4.77, p < .05). Consistent with hypothesis 2, sumer attributions for performance to the self (with no in-
state self-esteem was higher (Mcontrol ¼ 4.35, SD ¼ 1.46 vs. crease in attributions to the brand). Thus together, studies
Mstrong ¼ 4.93, SD ¼ 1.16; F(1, 89) ¼ 4.39, p < .05) and 1, 2, and 3 provide support for hypotheses 1, 2, and 3: a
anxiety was lower (Mcontrol ¼ 3.40, SD ¼ 1.85 vs. positive brand placebo effect on objective performance
Mstrong ¼ 2.73, SD ¼ 1.45; F(1, 89) ¼ 3.74, p ¼ .056) for that is attributed to the self and is mediated by state self-es-
the strong performance brand; figure 2 offers an teem enhancement and anxiety reduction.
illustration. In the subsequent studies, we explore boundary conditions
that alter the performance brand placebo effect on objective
Serial Mediation. What role did state self-esteem and
outcomes. Study 4 focuses on the moderating role of stress
anxiety play in improving objective performance? To as-
mindset (Crum et al. 2013), thereby shedding further light on
sess mediation, we conducted a bootstrapping analysis
the role of anxiety in determining the performance brand pla-
(Preacher and Hayes 2004, 2008; Zhao et al. 2010) with
cebo. Study 5 focuses on the moderating role of domain self-
performance brand condition as the independent variable,
efficacy beliefs (Bandura 1997, 2006), and also distinguishes
state self-esteem and anxiety as serial mediators, and ob-
performance brands from other brands that are highly re-
jective performance as the dependent variable. The pre-
garded by consumers (i.e., prestige brands; Broniarczyk and
dicted indirect effect via state self-esteem and, in turn,
Alba 1994; Park, Milberg, and Lawson 1991).
anxiety was significant (Indirect effect ¼ .036, 95% CI,
003–.131), supporting hypothesis 2. That is, a strong per-
formance brand heightened state self-esteem, which in turn STUDY 4: STRESS MINDSET AND
decreased anxiety and thereby improved objective perfor- REVERSAL OF THE PERFORMANCE
mance. Analyses do not support mediation via motivation BRAND PLACEBO
or enjoyment; a separate bootstrapping mediation model
examining the indirect effects of anxiety, motivation, and The primary objective of study 4 is to provide further ev-
enjoyment in parallel revealed that the anxiety pathway re- idence for the role of anxiety reduction in the performance
mained marginally significant (Indirect effect ¼ .068; 90% brand placebo via a theoretically relevant moderator and940 JOURNAL OF CONSUMER RESEARCH
FIGURE 2
COGNITIVE PERFORMANCE BRAND PLACEBO (STUDY 3)
Panel A: Objective Performance Panel B: State Self-Esteem Panel C: Task Anxiety
Downloaded from https://academic.oup.com/jcr/article-abstract/42/6/931/2358047 by UNIVERSITY OF THE ARTS LONDON user on 28 February 2019
(Correct test questions out of 5) (1-7 Scale) (1-7 Scale)
5 7 7
6 6
4
5 5
3 4 4
3 3
2
2 2
1 1 1
Control Performance Control Performance Control Performance
Brand Brand Brand
boundary condition. Study 3 provides support for the role enhancing increase, the placebo effect should weaken and
of anxiety by showing how a strong performance brand re- may reverse; that is, performance may worsen because the
duces anxiety and therefore enhances performance. placebo alleviates stress. Formally,
Because performance contexts can lead to stress that cre-
ates anxiety, the present study investigates individual be- H4: The positive impact of the performance brand placebo
increases (decreases) as the belief that stress is debilitating
liefs about the nature of experienced stress (stress enhances
(enhancing) increases.
vs. stress debilitates; Crum et al. 2013) and their impact on
the performance placebo. That is, we provide further pro-
If supported, the present study will (1) demonstrate a
cess evidence for the underlying anxiety mechanism via
boundary condition that weakens or even reverses the perfor-
the theoretically relevant moderator of stress mindset (cf.
mance brand placebo, and (2) further support the role of anx-
Spencer, Zanna, and Fong 2005).
iety as the underlying mechanism for the performance brand
Our identification of the role of anxiety in determining
placebo. Further, this study will also (3) demonstrate that
the performance brand placebo introduces stress mindset as
stress mindset influences actual task performance outcomes,
a mechanism by which this effect may potentially be
an effect that Crum et al. (2013) were unable to document.
heightened, lessened, or even reversed. Recent research
As a secondary objective, the current study also provides
has shown that exposure to environmental stressors (con-
further evidence for generalizability of a performance
sistent with those that induce anxiety) may have varying
brand placebo on cognitive performance. Whereas studies
effects on individuals based on individual stress mindset
1 to 3 manipulated the brand of a tangible good, this study
(Crum et al. 2013). Specifically, a majority of individuals
investigates service brands positioned on performance—
hold the belief that stress is debilitating: the experience of
namely, cognitive test preparation (such as Kaplan and
stress typically undermines self-reported psychological
Princeton Review) brands—and their impact on cognitive
health and work performance, consistent with the concep-
test performance.
tualization of stress as inducing maladaptive anxiety.
However, a minority of individuals believe that stress is en-
hancing: rather than maladaptive anxiety, stress actually Method
improves self-reported psychological health and work per- Participants and Design. The experimental design was
formance. This view is consistent with recent research that a 2 group (strong performance brand/weak performance
indicates that some individuals may reframe anxiety as ex- brand) between-subjects design, with a continuous measure
citement and thereby enhance performance outcomes of stress mindset (Crum et al. 2013). Participants were 84
(Brooks 2013). We have previously argued that a decrease students (54% male) at a large midwestern university who
in performance-induced anxiety underlies the performance participated voluntarily in return for course credit.
brand placebo. If so, then the strength of the positive pla-
cebo should increase with personal beliefs that stress is de- Procedure. Participants first completed an established
bilitating. In contrast, as personal beliefs that stress is measure of stress mindset (Crum et al. 2013). The stressGARVEY ET AL 941
mindset scale measures the degree to which individuals be- function of brand condition, stress mindset (a ¼ .78;
lieve that stress enhances versus debilitates performance. M ¼ 3.54, SD ¼ .87; mean centered), and their interaction.
Sample items include “Experiencing stress enhances my Analysis of variance revealed the expected two-way interac-
performance and productivity” and “The effects of stress tion (F(1, 80) ¼ 11.35, p ¼ .001); main effects were not sig-
Downloaded from https://academic.oup.com/jcr/article-abstract/42/6/931/2358047 by UNIVERSITY OF THE ARTS LONDON user on 28 February 2019
are negative and should be avoided” (measured on 7 point nificant (F’s < 1). We note that between manipulated
scales with end points “Strongly disagree” to “Strongly conditions, stress mindset did not vary significantly (F < 1);
agree”; the online appendix shows the full 8 item scale). nor did the time spent on the lesson or questions (F’s < 1.5).
The measure of stress mindset was embedded within a lon- To understand the nature of the interaction, spotlight
ger questionnaire to disguise its purpose. analyses were conducted at higher and lower levels of
In an ostensibly unrelated task, participants next were in- stress mindset (61 SD). A significant positive effect
formed that they would be trialing a new test preparation emerged at low levels of stress mindset (i.e., stress was
smartphone application that delivers lessons to improve seen as strongly debilitating) (b ¼ .91; t(80) ¼ 2.22,
performance on the Graduate Management Admission Test p < .05), whereas this effect reversed at high levels of
(GMAT). Based on a pretest (described later), participants stress mindset (i.e., stress was seen as strongly enhancing)
in the strong performance brand condition were told that (b ¼ .1.05; t(80) ¼ 2.56, p < .05). See figure 3 for an il-
the developer was Kaplan (a well-established brand famil- lustration. Floodlight analysis results (Johnson and Fay
iar to participants), whereas those in the weak performance 1950; Spiller et al. 2013) were also consistent with our the-
brand condition were told that the developer was Laserprep ory: A strong performance brand has an enhancing effect at
(a fictitious brand unknown to participants). Aside from stress mindset levels below the Johnson-Neyman point of
the brand name, the introduction did not differ by condi- 2.81 (t(80) ¼ 1.99, p ¼ .05) and a debilitating effect at
tion; the full text is available in the online appendix. stress mindset levels above 4.09 (t(80) ¼ 1.99, p ¼ .05).
To give participants actual consumption experience with Consistent with hypothesis 4, a strong performance brand
the performance brand, participants then sampled a lesson de- improved objective performance when stress was seen as
scribed as taken from the target test prep module. All partici- debilitating but undermined performance when stress was
pants proceeded through the same lesson, adapted from seen as enhancing.
Kaplan (2004, 41). The lesson took approximately 5 minutes
and provided material regarding how to answer critical rea-
soning questions. After completing the lesson, participants Discussion
answered five critical reasoning multiple-choice questions
Study 4 provides a deeper understanding of the role of
taken from Kaplan (2004)’s GMAT practice question bank (a
anxiety in determining the strength and direction of the pla-
sample item is shown in the online appendix). Finally, partici-
cebo effect. Specifically, whether branded performance
pants responded to background questions (e.g., gender, age).
Results FIGURE 3
Pretest. A pretest was administered to 57 students on a
COGNITIVE PERFORMANCE BRAND PLACEBO AS A
large midwestern university campus to rate how lessons FUNCTION OF STRESS MINDSET (STUDY 4)
from the two test preparation companies (i.e., strong per-
formance brand Kaplan and weak performance brand 5
Laserprep) were expected to influence GMAT perfor- Weak Performance Brand
(Correct test questions out of 5)
mance. As before, we used the following three items to 4 Strong Performance Brand
Objective Performance
measure brand performance expectations (all on a scale
from 1 [Not at all] to 7 [Very much]; adapted from Shiv
et al. 2005): “The lesson from [brand] will harm/help my 3
GMAT performance”; “I feel that the lesson from [brand]
is very bad/very good at improving my GMAT perfor- 2
mance”; “To what extent could the lesson from [brand] im-
prove your GMAT performance?” Performance
expectancies (a ¼ .96) were significantly higher for the 1
strong performance brand than for the weak performance
brand (Mstrong ¼ 4.63, SD ¼ 1.04 vs. Mweak ¼ 3.85, 0
SD ¼ 1.15; F(1, 55) ¼ 11.39, p < .01). Stress Mindset: Stress Mindset:
Debilitates (-1 SD) Enhances (+1 SD)
Objective Performance. Analysis of performance (num-
ber of test questions correctly answered) was conducted as aYou can also read
