Content uploaded by Joseph Cesario
Author content
All content in this area was uploaded by Joseph Cesario on Oct 21, 2017
Content may be subject to copyright.
Running Head: POWER POSING AND DYADIC INTERACTIONS 1
in press, Social Psychological and Personality Science
Pre-copyediting version -- please do not quote directly
Power Poseur:
Bodily Expansiveness Does Not Matter in Dyadic Interactions
Joseph Cesario and David J. Johnson
Michigan State University
Author Note
Joseph Cesario and David J. Johnson, Psychology Department, Michigan State
University.
All data and materials, including experimenter scripts, are available on the first author’s
website (www.cesariolab.com). We report all studies conducted in our lab testing these
hypotheses. The material is based upon work supported by National Science Foundation under
award No. BCS-1230281 to the first author. Portions of this manuscript were presented at the
Duck Conference on Social Cognition (2013).
Correspondence should be addressed to Joseph Cesario, Psychology Building, 316
Physics Road - Room 255, East Lansing, MI 48824 cesario@msu.edu
Running Head: POWER POSING AND DYADIC INTERACTIONS 2
Abstract
Strong claims have been made that power poses can significantly improve one's life.
Starting from an evolutionary perspective, we reason that expansive poses will have no impact in
more realistic situations, as in the presence of an interaction partner or when participants are
aware of what the pose should accomplish. Across four dyadic studies including both commonly-
used outcomes as well as a negotiation task (which could actually have direct benefits for one's
life), we find nearly uniform null effects of holding expansive poses, despite checks confirming
the success of the manipulation. For example, in two of the studies, participants watched a
popular TED talk on power poses, held an expansive pose, and then completed a negotiation in
the presence of a partner, as might happen in real life. We argue that researchers should stop
recommending power poses as an empirically-supported strategy for improving one's life.
Keywords: power poses, power, evolutionary psychology
Running Head: POWER POSING AND DYADIC INTERACTIONS 3
Power Poseur:
Bodily Expansiveness Does Not Matter in Dyadic Interactions
Can making one’s body physically expansive influence the psychology of power and
have positive effects on life outcomes? The possibility that “power poses” can have cognitive,
physiological, and behavioral effects has important theoretical and practical implications. On the
theoretical side, this research has been used to support an embodied (rather than computational)
understanding of the mind. On the practical side, power posing has been offered as an easy
technique for powerless people to achieve the outcomes they want in life. As one indicator of
how this research has captured public attention, over 40 million people have viewed an online
TED talk recommending power poses (Cuddy, 2012). It is difficult to think of many topics in
social psychology that have so quickly captured such wide public interest. In this manuscript, we
critique the theoretical foundation of this research and test its practical claims.
Past Research and Theory
Past research has shown that holding expansive bodily poses (compared to constrictive
poses) can positively impact a range of power-related outcomes. (See Figure 1 for examples.)
Holding expansive poses can increase the subjective experience of power (Carney, Cuddy, &
Yap, 2010; Huang, Galinsky, Gruenfeld, & Guillory, 2011), risk-taking behavior (Carney et al.,
2010; Huang et al., 2011), abstract thinking (Huang et al., 2011), testosterone (Carney et al.,
2010), the implicit activation of power (through a word completion task; Huang et al., 2011),
pain tolerance (Bohns & Wiltermuth, 2012), and effectiveness at mock job interviews (Cuddy,
Wilmuth, Yap, & Carney, 2015).
Running Head: POWER POSING AND DYADIC INTERACTIONS 4
Figure 1. Constrictive (top panel) and expansive (bottom panel) poses used in past
research and the present studies.
Researchers have combined ideas from evolutionary theories and the "embodiment"
literature to argue that expansive poses have a direct and unmediated effect on the psychology of
power. The argument is that because size and power have been closely tied throughout evolution,
power is "embodied" and therefore increasing one’s size should induce power. Carney (2010) et
al. provide a prototypical description of this mechanism: “Humans and other animals display
power and dominance through expansive non-verbal displays, and these power poses are deeply
intertwined with the evolutionary selection of what is ‘alpha’ (p. 1363)." Along with this is the
claim that the connection between physical pose and power is “deeply intimate” and automatic.
For instance, Adam and Galinsky (2012) state “In embodied cognition, the link between a
physical experience and its symbolic meaning is direct, as it is the physical experience itself that
Running Head: POWER POSING AND DYADIC INTERACTIONS 5
carries the symbolic meaning. In other words, the symbolic meaning is always automatically
embodied because it directly stems from the physical experience” (p. 919).
A Critical Look at Existing Theory
Although power poses research might be consistent with an embodiment approach (but
see Wilson & Golonka, 2013), we question whether this account is adequate given current
evolutionary thought. Previous uses of evolutionary logic may reflect a misunderstanding of the
role of physical size in animal contests (see also Gawronski & Cesario, 2013). A detailed
description of game theory and the logic of animal contests is beyond this manuscript (see
Maynard-Smith, 1974, 1979; Maynard-Smith & Parker, 1976; Maynard-Smith, & Price, 1973). It
is sufficient here to note that physical size is not the only variable used by the brain when
deciding which actions to pursue. Research on animal contests has shown animals take into
account: relative judgments of the other animal’s physical size, coalitional support,
environmental constraints, positions in social hierarchies, and past encounters with the target
(see, e.g., Benson-Amram, Heinen, Dryer, & Holekamp, 2011; D.C. Blanchard, 1997; D.C.
Blanchard & Blanchard, 1984; D.C. Blanchard, Hynd, Minke, Minemoto, & Blanchard, 2001;
R.J. Blanchard, Flannelly, & Blanchard, 1986; McComb, Packer, & Pusey, 1994; Smith et al.,
2010; Stankowich & Blumstein, 2005; Wilson, Britton, & Frank, 2002).
The idea that power can be increased by a temporary expansion of one's physical size --
overriding other relevant role and context-related information -- is akin to arguing that the brain
can be “tricked” into regulating the body to such a degree as to result in behaviors that would
negatively impact reproductive success. Any animal who escalated conflict against others with
better fighting skills, weaponry, or status -- just because of momentary expansiveness -- would
have experienced drastically-reduced reproductive success, making this an unlikely candidate for
Running Head: POWER POSING AND DYADIC INTERACTIONS 6
an evolutionary adaptation. Rather, the role of physical size in animal contests is computational
and comparative, not a direct mapping of physical size and power.
In fact, the relevant animal research does not instill confidence that power poses could
successfully “boost” one’s power. Physical size, as it relates to one’s position in a dominance
hierarchy, is an honest signal in that it reveals actual qualities about an individual’s ability to
attain and maintain hierarchy positions. In this context, power poses must be understood as
elevating a person’s internal representation of his or her status beyond that which is justified by
the person’s actual status (as when the poor or powerless hold expansive poses). Hence power
poses fall within the class of behavior known as dishonest signals. When dishonest signals are
easy to produce (as in raising one’s arms), their use by cheaters is selected against through other
kinds of costs, including challenges by those who possess the signals honestly.
For example, the comb of male red junglefowl (Gallus gallus) is affected by social status
and serves as an honest signal of mate quality. High-status males aggressively challenge other
males who display large combs and dominant behavior, thereby preventing dishonest
development of combs by subordinates (Parker & Ligon, 2002). In one experiment, when
subordinate male Harris Sparrows (Zonotrichia querula) had the feathers on their throats and
crowns artificially dyed a darker black (a signal of mate quality), they were persecuted more
strongly by high status males (Rohwer, 1977; Rohwer & Rohwer, 1978).1 In these species, trying
to “fake it ‘till you become it" (as recommended by Cuddy, 2012) does not result in better
outcomes. The implication for the low-status worker, entering the boss's office having spent two
minutes with a puffed-out chest, should be clear.
Why Does It Matter?
1 We must quote Rohwer (1977): “Shortly after the dyeings it was evident that the manipulation had precipitated a
veritable wrath of aggression from the legitimate studlies” (p. 114, emphasis added).
Running Head: POWER POSING AND DYADIC INTERACTIONS 7
If power-related decisions are governed not by direct feedback but by a computational
system that takes into account broader features of an interaction, there may be serious
implications for the practical utility of power pose interventions. There are two reasons for
concern.
First, studying any computational process requires ensuring that the relevant pieces of
information used by real decision-makers are present in laboratory investigations. If power pose
studies show positive laboratory effects but are missing critical pieces of information used by
real-world decision-makers, then the real-world applicability of these findings may be limited. In
the case of displaying powerful behavior, evolutionary theory argues that this necessarily
includes partner information. Unfortunately, the prototypical power pose experiment involves a
single participant holding an expansive or constrictive pose, even though the intention is to have
people use power poses in dyadic situations (as when a pose is held prior to negotiating with the
boss). If there is any real-world value in power poses, experimental tests must include these
critical dimensions.2
Second, there is the issue of awareness. In order to enact power poses in the real-world, a
person has to know what the pose is supposed to do. In contrast, almost all existing research
keeps participants blind to the purpose of the pose (see Keller, Johnson, & Harder, in press;
Ranehill et al., 2015), and researchers go to great lengths to mislead participants. But this is an
impossible condition for the real world. If I am to change my outcomes by intentionally holding
a power pose before a job interview, I have to know that this is the intended effect. Indeed,
supporting the possibility that awareness matters, Carney, Cuddy, and Yap (2015) specifically
cite awareness as a moderator that could undermine expansive pose effects.
2 One underpowered study (Cuddy et al., 2015) had participants deliver a speech to two evaluators. However, a
higher-powered peer-reviewed preregistered replication failed to replicate these results (Keller, Johnson, &
Harder, in press).
Running Head: POWER POSING AND DYADIC INTERACTIONS 8
If there is a direct link between expansive body postures and power, then it makes sense
to recommend power poses to low-status individuals on the way to negotiate a raise with the
boss. But if other, more important information such as knowledge one's role or position in a
hierarchy also enters the computation (see Cesario & McDonald, 2013), we would be offering
false hope to powerless people with the message that their lives will drastically change by using
such techniques.
The Present Research
The first objective is to test the effects of power poses within the context of dyadic
interactions. Additionally, we include dependent measures from past research as well as a
negotiation task, an outcome that could have clear benefits for a person’s life. The second
objective is to test the effects of power poses with studies containing larger sample sizes and
identical dependent variables across studies. The tendency of past studies to use small samples
and report different dependent variables makes an overall assessment of power pose effects
difficult. (We conducted no a priori power analyses but instead aimed to collect as many
participants as possible each semester.) Across the studies, expansive poses were compared to
constrictive or neutral poses.3 Poses were those depicted in Figure 1 and instructions replicated
exactly those from the original Carney et al. (2010), with participants holding the two poses for 1
minute each. Table 1 presents descriptive statistics and Table 2 presents inferential statistics.
In Studies 1 and 2, participants watched the TED talk on power poses, held an expansive
pose, and then completed the dependent variables; responses between expansive pose and no
pose participants were compared. In Studies 3 and 4, participants held either an expansive or
constrictive pose alone (Study 3) or in the presence of another participant (Study 4), and then
3 Criteria for removal of participants were decided a priori: indicating made-up responses, indicating friendship
with their partner, or failing careless responding checks (e.g., "Select 'strongly agree' now").
Running Head: POWER POSING AND DYADIC INTERACTIONS 9
completed the dependent measures; responses between expansive and constrictive participants
were compared. In Studies 3 and 4, participants were blind to the purpose of holding the poses.
All participants were undergraduates at Michigan State University and received course credit for
participating.
Study 1: Holding the Pose After Watching a TED Talk
Procedure
For this study, we intended to make the experimental context as realistic as possible, in
terms of how a person might learn about and use power poses. Participants arrived in pairs.
Participants in the expansive pose condition watched the TED talk on power poses, after which
the experimenter said, “we’re interested in whether holding certain positions will increase your
power and make you more effective in negotiations.” The video was edited so that only the
introductory and concluding sections were presented (8 min. duration). No results were
described, as we did not want participants' behavior to be driven by demand effects. The
experimenter then instructed this participant to hold each of the two expansive poses.
While the "expansive participant" was watching the video and holding the poses, the "no
pose" participant was sitting alone in an adjacent room. This participant also watched the same
TED talk but was given no information about the purpose of the study or the experience of the
other participant. This participant was simply told “to watch a short video, and you’ll be asked
questions about it later on” and was asked to “sit comfortably with your hands on your legs
during and after the video.” (When probed, 22 participants revealed that they had spontaneously
held the poses described in the video and were removed from analyses, an a priori decision.)
Finally, participants were brought together and completed all measures.
Participants and Design
Running Head: POWER POSING AND DYADIC INTERACTIONS 10
Analyses compare the responses of participants who held the expansive pose (n=92) to
participants who held no pose (n=72).
Measures4
Dependent measures. Three dependent measures of power were collected: risk taking,
abstract thinking, and a negotiation task.
Risk-taking: Gambling task. This measure replicates risk-taking measures from past
power pose studies (Carney et al., 2010; Cesario & McDonald, 2013; Huang et al., 2011;
Ranehill et al., 2015). Participants were endowed with two tickets for a raffle to win a $50
grocery store gift card. They could either enter the two tickets right now (the safe bet) or take a
risk and role a die, potentially doubling their number of tickets (if a 4, 5, or 6 was rolled) but
potentially losing them all (if a 1, 2, or 3 was rolled). Participants’ choices were recorded. To
ensure decisions were independent both made their decisions confidentially in writing. Only at
the end of the study were these papers revealed and the dice rolled. To test whether expansive
poses increased gambling, Fisher's exact Odds Ratio tests compared the proportion of people
who chose the risky gamble between conditions.
Abstract Thinking: Gestalt task. Participants completed the Gestalt Completion Task
(Huang et al., 2011). Participants were shown images of incomplete pictures and asked to
identify them. We followed the exact scoring procedures used by Huang et al. and had
independent coders rate whether participants provided correct, incorrect subordinate (incorrectly
identifying a picture at a low abstraction level), or incorrect superordinate (incorrectly
identifying a picture at a high abstraction level) responses. These researchers obtained three pose
4 Participants also completed physical strength and attractiveness measures. Overall, stronger and more attractive
participants showed more powerful outcomes. Details are available from the first author. Unrelated dependent
variables were also collected for a different line of research on recalibration theory, as described in Johnson and
Cesario (in prep). The decision was made a priori that these measures would not be analyzed for the present
research.
Running Head: POWER POSING AND DYADIC INTERACTIONS 11
effects, all of which indicate that powerful people think more abstractly: 1) expansive poses
increased the number of pictures correctly identified; 2) expansive poses decreased the number
of subordinate-level incorrect responses; and 3) expansive poses increased the number of
superordinate-level incorrect responses. To test whether expansive poses increased abstract
thought, t-tests compared the means between conditions for each of the three calculations.
Negotiation task. Participants completed the “Synertech-Dosagen” negotiation task to
test whether holding expansive poses resulted in better negotiation outcomes (see Appelt, Zou,
Arora, & Higgins, 2009; Diekmann, Tenbrunsel, Shah, Schroth, & Bazerman, 1996; Galinsky,
Leonardelli, Okhuysen, & Mussweiler, 2005; Galinsky & Mussweiler, 2001; Kurtzberg, Naquin,
& Belkin, 2009). One participant is the seller and the other the buyer of a pharmaceutical plant;
participants receive identical general information plus information specific to their role.
Participants had 20 minutes to negotiate the sale with the explicit goal to get the best outcome for
themselves.
We recorded each participant’s initial first offer, which participant made the first offer,
and the final sale price (for those successful negotiations, which was nearly all dyads). These
outcomes have been used in prior research and there is “remarkably robust” (p. 1548, Gunia,
Swaab, Sivanathan, & Galinsky, 2013) evidence that setting an aggressive first offer yields better
final outcomes. As such, this is one of the most direct measures yet to assess an outcome that
could be life-changing in the way suggested by power pose advocates.
As negotiation outcomes are necessarily partner-dependent, we used a series of multilevel
models (with dyad as Level 2 and individual as Level 1) to test the effects of poses across the
different negotiation outcomes. To test whether expansive poses made participants more likely to
make the first offer, a multilevel model was conducted with negotiation role (-1=buyer, 1=seller)
Running Head: POWER POSING AND DYADIC INTERACTIONS 12
and pose condition (-1=neutral, 1=expansive) as predictors and whether that participant made the
opening bid (0=no, 1=yes) as the predicted variable. Evidence of the benefits of expansive poses
would be demonstrated with greater likelihood of making the opening bid for expansive pose
participants.
To test whether expansive poses made participants set a more aggressive opening offer, a
multilevel model tested for the interaction between negotiation role and pose, testing whether
sellers holding expansive (vs. neutral) poses set higher opening offers and buyers holding
expansive (vs. neutral) poses set lower opening offers.
The final test was whether holding expansive poses resulted in better final sale prices
(with higher prices for sellers and lower prices for buyers). Because outcome price is a single
outcome for each dyad, we created a value for each participant that represented the deviation of
the final sale price from the starting offer. This value represents success at the negotiation,
insofar as having a final price that was closer to one’s opening bid (i.e., a lower deviation value)
is better than having a final price far from one’s opening bid. The deviation score was calculated
such that smaller values were better for both sellers and buyers; hence a main effect of pose
condition would provide evidence for the benefits of expansive poses.5
To show that participants completed the task correctly, across studies the expected main
effect of negotiation role on opening bid was obtained, b=2.43, SE=0.16, t(281)=15.37, p < .001,
with sellers setting higher opening bids (M=23.50, SD=4.00) than buyers (M=18.78, SD=3.49).
In addition, setting an aggressive first offer strongly resulted in better final outcomes for both
buyers, r(136)=.37, p<.001, d=0.80, 95% CI=[0.43, 1.16], and sellers, r(127)=.76, p<.001,
5 Throughout the studies, a small number of dyads (n = 9) had offers suggesting they did not understand the
instructions or experimenter error. These dyads are removed from analyses, as this provides the most liberal test
of the effectiveness of expansive poses. Eight participants who made extreme first offers and were clear outliers
were also removed.
Running Head: POWER POSING AND DYADIC INTERACTIONS 13
d=2.34, 95% CI=[1.80, 2.88].
Manipulation Check. As a manipulation check, participants reported their felt power by
rating, on 1 to 7 scales, how powerful, in control, in charge, superior, dominant, afraid (r), at
risk (r), vulnerable (r), lowly (r), and subordinate (r) they currently felt, α=.79, 95% CI=[.75,
.83].6
Results
The manipulation check confirmed that the pose manipulation was successfully
implemented. Participants reported feeling more powerful after holding the expansive pose,
M=5.22, SD=0.68, compared to the neutral pose, M=4.92, SD=0.72, d=0.43, 95% CI=[0.12,
0.74]. As shown in Table 2, however, there were no significant effects of holding expansive
poses on any of the dependent measures.
6 In the original Carney et al. (2010), this was described as a dependent variable. In more recent publications, this
is described as a manipulation check (see Carney, Yap, Lucas, Mehta, McGee, & Wilmuth, in prep; Cuddy et al.,
2015). We are agnostic about this and we simply follow these authors in their latest terminology.
Running Head: POWER POSING AND DYADIC INTERACTIONS 14
Study 1
Study 2
Study 3
Study 4
Initial Sample Size
264
282
257
330
Removals For:
careless responding
n/a
n/a
24
34
made up responses
12
9
17
6
friends with partner
8
4
6
8
held pose in booth
22
16
n/a
n/a
Final Total N
226
254
213
284
N completed study with partner
164
161
140
247
N females, N males
122, 40
0, 161
123, 17
139, 108
N in key comparison conditions
164
161
71
174
N control, N expansive
72, 92
72, 89
34, 37
85, 89
% Risky Gamble:
control
expansive
59.2%
68.1%
80.3%
78.4%
47.1%
51.4%
60.0%
77.5%
Abstract Thought Task
Number correct:
control
expansive
3.56 (1.88)
3.71 (1.95)
3.63 (2.04)
3.76 (1.92)
3.47 (1.91)
3.27 (1.94)
3.31 (1.93)
3.43 (1.84)
Incorrect subordinate:
control
expansive
0.68 (0.82)
0.65 (0.94)
1.67 (1.87)
1.47 (1.67)
0.94 (1.10)
1.35 (2.04)
0.86 (1.35)
1.03 (1.47)
Incorrect superordinate:
control
expansive
0.29 (0.54)
0.29 (0.55)
0.04 (0.20)
0.11 (0.51)
0.21 (0.41)
0.41 (0.69)
0.27 (0.54)
0.31 (0.54)
Negotiation Task
% first offer:
control
expansive
45.7%
53.9%
58.0%
44.2%
50.0%
54.3%
n/a
Sellers' first offer amount:
control
expansive
27.91 (17.72)
28.59 (22.37)
26.18 (13.64)
23.30 (4.45)
22.73 (3.56)
23.25 (10.88)
n/a
Buyers' first offer amount:
control
expansive
19.40 (3.75)
18.39 (3.47)
19.14 (3.59)
18.36 (3.24)
17.95 (3.59)
17.63 (4.26)
n/a
Final price deviation:
control
expansive
3.54 (7.66)
3.16 (7.80)
2.76 (2.31)
3.21 (9.37)
4.46 (11.37)
3.24 (4.49)
n/a
Running Head: POWER POSING AND DYADIC INTERACTIONS 15
Table 1. Descriptive statistics across the studies; values in parentheses are standard
deviations. Descriptives are provided on the subset of data reported in analyses in the text.
"Control" condition refers to holding the neutral pose (Studies 1 and 2) or the constrictive pose
(Studies 3-4). Study 4 has no negotiation data because this task was not included. "N completed
study with partner" refers to the number of participants who had a partner present during the
study; "N in key comparison conditions" is the number of expansive vs. control condition
participants and is the total on which analyses are conducted.
Running Head: POWER POSING AND DYADIC INTERACTIONS 16
Study1
Study 2
Study 3
Study 4
Risky Gamble:
OR = 0.68
[0.34, 1.36],
d = 0.21
[-0.14, 0.57]
OR = 1.12
[0.48, 2.64],
d = -0.06
[-0.49, 0.37]
OR = 0.84
[.30, 2.36],
d = 0.09
[-0.43, 0.62]
OR = 0.44
[0.21, 0.88],
d = 0.46
[0.09, 0.83]
Abstract Thought Task
Number correct:
d = 0.08
[-0.23, 0.39]
d = 0.06
[-0.25, 0.37]
d = -0.10
[-0.57, 0.36]
d = 0.07
[-0.23, 0.36]
Incorrect subordinate:
d = 0.03
[-0.28, 0.34]
d = 0.11
[-0.20, 0.42]
d = -0.25
[-0.71, 0.22]
d = -0.12
[-0.42, 0.17]
Incorrect superordinate:
d = 0.003
[-0.31, 0.31]
d = 0.18
[-0.14, 0.49]
d = 0.35
[-0.12, 0.82]
d = 0.08
[-0.22, 0.38]
Negotiation Task
% first offer:
b = 0.17, SE =
0.16, z(159) =
1.04
OR = 1.39,
d = 0.09
[-0.08, 0.27]
b = -0.33, SE =
0.17, z(155) =
-1.89,
OR = 0.58,
d = -0.18
[-0.37, 0.002]
b = 0.14, SE =
0.25, z(69) =
0.58,
OR = 1.18,
d = 0.08 [-
0.19, 0.35]
n/a
First offer amount:
b = 0.42, SE =
1.26,
t(143) = 0.33
b = -054, SE =
0.64,
t(145) = -0.84
b = 0.21, SE =
0.84,
t(58) = 0.25
n/a
Seller's first offer amount:
d = 0.03
[-0.45, 0.51]
d = -0.29
[-0.77, 0.19]
d = 0.06
[-0.75, 0.87]
n/a
Buyer's first offer amount:
d = 0.28
[-0.20, 0.76]
d = 0.23
[-0.24, 0.70]
d = 0.08
[-0.62, 0.79]
n/a
Final price deviation:
b = -0.27, SE =
0.65,
t(134)= -0.42,
d = 0.05
[-0.29, 0.39]
b = 0.18, SE =
0.61,
t(137)= 0.30,
d = -0.06
[-0.40, 0.27]
b = -0.78, SE =
1.16,
t(55)= -0.67,
d = 0.14
[-0.39, 0.67]
n/a
Table 2. Analyses testing differences between expansive pose conditions and neutral pose
(Studies 1 and 2) or constrictive pose (Studies 3-4) conditions. d is Cohen's d; numbers in
bracket are 95% CIs.
Running Head: POWER POSING AND DYADIC INTERACTIONS 17
Study 2: Holding the Pose After Watching an Instructional TED Talk, Males Only
Study 2 was a direct replication of Study 1, except that we recruited only males given the
small percentage of males in Study 1. Analyses compare the responses of participants who held
the expansive pose (n=89) to participants who held no pose (n=72).
Results
The manipulation check showed the predicted directional effect, such that participants
felt slightly more powerful after holding the expansive pose, M=5.19, SD=0.67, compared to the
neutral pose, M=5.02, SD=0.81, d=0.24, 95% CI=[-0.08, 0.55].) As with the prior study, and
shown in Table 2, participants in the expansive condition showed no evidence of greater power
on any dependent measure.
Study 3: Holding the Pose Without Awareness
Procedures
In this study, several changes were made to make the method more similar to past power
pose research. One participant in the dyad held either an expansive or constrictive pose while the
other dyadic partner waited alone, in another room, until the participant finished holding the
poses. As in past research, the participant who held the expansive or constrictive pose had no
awareness of why he or she was holding the pose, as a cover story kept the participant blind to
the purpose of the study. The partner returned to the room and the pose participants completed
all dependent measures in the presence of the other participant.
We also tried to make the experimental situation match real world circumstances by
instructing the participant holding the pose:
“While you’re holding this position, I’d like you to imagine that in a few minutes
the other participant is going to be sitting in that chair and that you’re going to
Running Head: POWER POSING AND DYADIC INTERACTIONS 18
compete against him/her in a negotiation, and you’re going to try to get as much
for yourself out of the negotiation as you can.”
This mimics the advice given to the public by power pose researchers: Immediately
before a negotiation, job interview, etc., hold the pose and think about the upcoming event.
Participants and Design
Analyses compare the responses of participants who held the expansive pose (n=37) to
those who held the constrictive pose (n=34), given that this is the most common comparison in
past research.
Results
Again, the manipulation check indicated that participants felt more powerful after holding
the expansive pose, M=4.84, SD=0.64, than the constrictive pose, M=4.22, SD=0.85, d=0.83,
95% CI=[0.34, 1.31]. As seen in Table 2, holding the expansive pose had no statistically-
significant effects on any measured outcome.7
Study 4: Holding the Pose in the Presence of a Partner
Procedure
In this study, participants held either an expansive or constrictive pose across the table
from their partner, who held a neutral pose. The participant holding the expansive or constrictive
pose was first instructed on how to hold the pose. The neutral pose participant was then
instructed to sit in a neutral position, with arms on the arm rests and a straight back. While power
pose researchers explicitly recommend against holding poses in front of others (e.g., Cuddy,
2012), we suspected from past work on complementarity that this might actually have stronger
effects as expansive and constrictive poses signal dominance and submissiveness within dyadic
7 Because only one participant in each dyad held a pose and this study compares expansive vs. constrictive
conditions, multilevel models were not needed for the negotiation analyses.
Running Head: POWER POSING AND DYADIC INTERACTIONS 19
interactions (e.g., Tiedens & Fragale, 2003).
Participants and Design
Analyses compare the responses of participants who held the expansive pose (n=89) to
those who held the constrictive pose (n=85). Participants were always in same-sex dyads, with
male participants paired with male confederates but female participants paired with other female
participants.8
Results
The manipulation check showed that participants reported feeling more powerful after
holding the expansive pose, M=4.77, SD=0.78 than the constrictive pose, M=4.37, SD=0.82,
d=0.49, 95% CI=[0.19, 0.80]. The only effect to reach significance in this study was the effect on
the risky gamble, d=.46, 95% CI=[.09, .83]. As seen in Table 2, no other effects reached
statistical significance.
General Discussion
Across four studies using dyadic designs, we found no evidence that power poses had
beneficial effects on power-related outcomes. These studies used more realistic methodologies
and, in total, had a larger sample size than any previous study on this topic. Furthermore, our
three largest studies were larger than 91% of past studies finding effects of expansive poses (see
Carney et al., 2015, Table 1). Most important, Studies 1 and 2 recreated the conditions under
which a person might encounter power pose recommendations.
It is common to acknowledge that any study is simply a datapoint for a future meta-
analysis, and our studies of course are no exception. However, one must balance this fact with
past failures to replicate, small samples in publications that have obtained significant results,
8 Dyadic composition was different in this study to test predictions for other research regarding the effects of
physical strength and attractiveness.
Running Head: POWER POSING AND DYADIC INTERACTIONS 20
little evidence of moderators, and p-curve analyses that cast doubt on the published findings
(Simmons & Simonsohn, 2015). That is, in additional to the published data, the broader research
and publication details surrounding those data are relevant information. Given the scope and size
of the four studies presented here, these data should give pause to those who recommend power
poses as a means for the poor and low-status to better their lives.
Across all studies, the most consistent pose effect was on felt power. While power pose
researchers consider this a manipulation check and have in the past explicitly downplayed the
importance of this measure (Cuddy et al., 2015), defenders could argue this effect is important
for actual life outcomes. If you can “fake it ‘till you become it,” then feeling more powerful
might give you the motivation to carry through initial failures until you succeed. At the same
time one could easily predict the opposite: Failing repeatedly when you expect to succeed
(because you held expansive poses and felt powerful) may decrease motivation and increase
uncertainty regarding the connection between one’s actions and outcomes. Indeed, one might
expect this negative outcome to be most likely for low status individuals, who are more likely to
lack the skills needed for success. Those who believe power poses are an effective method
empowerment must provide concrete evidence that the poses have effects beyond just increased
feelings of power.
Limitations
Before discussing limitations of the current research, it is important to note that our
failure to find effects of power poses should not be attributed experimenter or participant
incompetence. Manipulation checks were successful, and participants showed predictable
patterns of bargaining in the negotiations.
Nonetheless, there are limitations of the current work. All data come from a single
Running Head: POWER POSING AND DYADIC INTERACTIONS 21
population, undergraduates at Michigan State University. This might be a problem if, for
example, these students were not skilled negotiators. Perhaps participants were so unskilled that
even power poses could not impact their performance. Much existing work on power poses
involves business school students (e.g., Cuddy et al., 2015), who may be meaningfully different
than the general undergraduate population. However, this interpretation undermines the entire
premise of power poses, as they are something that everyday people and low status individuals
can use to better their outcomes.
Another limitation of Studies 1 and 2 should be acknowledged. In these studies, the
participant who did not hold a pose did watch the TED talk video. It is possible that merely
watching this video had some benefit for the participant (e.g., increased confidence) even if that
participant did not hold the pose him or herself. Perhaps this benefit of seeing the video worked
against obtaining positive behavioral effects of holding an expansive pose. We do not believe
this design choice compromised our findings for three reasons. First, participants who held the
expansive pose did report more felt power than those who merely watched the video, meaning
that the manipulation check did show that holding the pose had its intended effect. Thus merely
watching the video was not identical to watching the video and holding the expansive pose.
Second, we suspect that power pose advocates would not actually want to raise this argument, as
it would undermine the premise that holding expansive poses is important. Finally, Studies 3 and
4, which were more closely aligned with the traditional methodology used in power pose
research, also failed to find any positive effect of expansive poses.
Conclusion
There has been unprecedented public interest in the possibility that holding expansive
poses can yield better life outcomes, particularly for powerless and low-status individuals. The
Running Head: POWER POSING AND DYADIC INTERACTIONS 22
current work provides two major arguments for questioning this narrative. First, the theoretical
underpinning is unlikely in light of a more precise understanding of evolution. Second, more
realistic conditions yielded no impact on power-related outcomes. In light of these points, as well
as past criticisms and failures to replicate pose effects, we suggest ceasing to recommend this
technique to the low-status and powerless until more supportive data can be gathered.
Running Head: POWER POSING AND DYADIC INTERACTIONS 23
References
Adam, H., & Galinsky, A.D. (2012). Enclothed cognition. Journal of Experimental Social
Psychology, 48(4), 918–925. doi:10.1016/j.jesp.2012.02.008
Appelt, K.C., Zou, X., Arora, P., & Higgins, E.T. (2009). Regulatory fit in negotiation: Effects of
“prevention-buyer” and “promotion-seller” fit. Social Cognition, 27(3), 365–384.
Benson-Amram, S., Heinen, V.K., Dryer, S. L., & Holekamp, K.E. (2011). Numerical
assessment and individual call discrimination by wild spotted hyaenas, Crocuta crocuta.
Animal Behaviour, 82, 743-752.
Blanchard, D.C. (1997). Stimulus and environmental control of defensive behaviors. In M.
Bouton, & M. Fanselow (Eds.), The functional behaviorism of Robert C. Bolles:
Learning, motivation and cognition (p. 283 - 305). Washington, DC: American
Psychological Association.
Blanchard, D.C., & Blanchard, R.J. (1984). Affect and aggression: An animal model applied to
human behavior. In R.J. Blanchard & D.C. Blanchard (Eds.), Advances in the study of
aggression (Vol. 1, pp. 2-62). New York, NY: Academic Press.
Blanchard, D.C., Hynd, A.L., Minke, K.A., Minemoto, T., & Blanchard, R.J. (2001). Human
defensive behaviors to threat scenarios show parallels to fear- and anxiety-related defense
patterns of non-human animals. Neuroscience and Biobehavioral Reviews, 25, 761-770.
Blanchard, R.J., Flannelly, K.J., & Blanchard, D.C. (1986). Defensive behaviors of laboratory
and wild Rattus norvegicus. Journal of Comparative Psychology, 100, 101-107.
Bohns, V.K., & Wiltermuth, S.S. (2012). It hurts when I do this (or you do that): Posture and
pain tolerance. Journal of Experimental Social Psychology, 48(1), 341–345.
doi:10.1016/j.jesp.2011.05.022
Running Head: POWER POSING AND DYADIC INTERACTIONS 24
Carney, D.R., Cuddy, A.J.C., & Yap, A.J. (2010). Power posing: brief nonverbal displays affect
neuroendocrine levels and risk tolerance. Psychological Science, 21(10), 1363–8.
doi:10.1177/0956797610383437
Carney, D.R., Cuddy, A.J.C., & Yap, A.J. (2015). Review and summary of research on the
embodied effects of expansive (vs. contractive) nonverbal displays. Psychological
Science. Advance online publication. DOI: 10.1177/0956797614566855
Carney, D.R., Yap, A.J., Lucas, B.J., Mehta, P.H., McGee, J., & Wilmuth, C. (in prep). Power
buffers stress – for better and for worse.
Cesario, J., & McDonald, M.M. (2013). Bodies in context: Power poses as a computation of
action possibility. Social Cognition, 31(2), 260–274. doi:10.1521/soco.2013.31.2.260
Cuddy, A.J.C. (2012). Your body language shapes who you are. Retrieved from
http://www.ted.com/talks/amy_cuddy_your_body_language_shapes_who_you_are?c=91
3722
Cuddy, A.J.C., Wilmuth, C.A., Yap, A.J., & Carney, D.R. (2015). Preparatory power posing
affects nonverbal presence and job interview performance. Journal of Applied
Psychology. Advance online publication. http://dx.doi.org/10.1037/a0038543
Diekmann, K.A., Tenbrunsel, A.E., Shah, P.P., Schroth, H.A., & Bazerman, M.H. (1996). The
descriptive and prescriptive use of previous purchase price in negotiations.
Organizational Behavior and Human Decision Processes, 66(2), 179–191.
Galinsky, A.D., Leonardelli, G.J., Okhuysen, G.A, & Mussweiler, T. (2005). Regulatory focus at
the bargaining table: Promoting distributive and integrative success. Personality & Social
Psychology Bulletin, 31(8), 1087–98. doi:10.1177/0146167205276429
Galinsky, A.D., & Mussweiler, T. (2001). First offers as anchors: The role of perspective-taking
Running Head: POWER POSING AND DYADIC INTERACTIONS 25
and negotiator focus. Journal of Personality and Social Psychology, 81(4), 657–669.
doi:10.1037//0022-3514.81.4.657
Gunia, B.C., Swaab, R.I., Sivanathan, N., & Galinsky, A.D. (2013). The remarkable robustness
of the first-offer effect: Across culture, power, and issues. Personality and Social
Psychology Bulletin, 39(12), 1547–1558. doi:10.1177/0146167213499236
Huang, L., Galinsky, A.D., Gruenfeld, D.H., & Guillory, L.E. (2011). Powerful postures versus
powerful roles: Which is the proximate correlate of thought and behavior? Psychological
Science, 22(1), 95–102. doi:10.1177/0956797610391912
Keller, V.N., Johnson, D.J., & Harder, J.A. (in press). Meeting your inner Super(wo)man: Are
power poses effective when taught? Comprehensive Results in Social Psychology.
Kurtzberg, T.R., Naquin, C.E., & Belkin, L.Y. (2009). Humor as a relationship-building tool in
online negotiations. International Journal of Conflict Management, 20(4), 377–397.
doi:10.1108/10444060910991075
Maynard-Smith, J. (1974). The theory of games and the evolution of animal conflicts. Journal of
Theoretical Biology, 47(1), 209–21.
Maynard-Smith, J. (1979). Game theory and the evolution of behavior. Proceedings of the Royal
Society B, 205, 475-488.
Maynard-Smith, J., & Parker, G.A. (1976). The logic of asymmetric contests. Animal Behaviour,
24, 159-175.
Maynard-Smith, J., & Price, G.R. (1973). The logic of animal conflict. Nature, 246, 15-18.
McComb, K., Packer, C., & Pusey, A. (1994). Roaring and numerical assessment in contests
between groups of female lions, Panthera leo. Animal Behaviour, 47, 379-387.
Parker, T.H., & Ligon, J.D. (2002). Dominant male red junglefowl (Gallus gallus) test the
Running Head: POWER POSING AND DYADIC INTERACTIONS 26
dominance status of other males. Behavioral Ecology and Sociobiology, 53, 20–24.
doi:10.1007/s00265-002-0544-5
Ranehill, E., Dreber, A., Johannesson, M., Leiberg, S., Sul, S., & Weber, R.A. (2015). Assessing
the robustness of power posing: No effect on hormones and risk tolerance in a large
sample of men and women. Psychological Science, 26, 653–656.
doi:10.1177/0956797614553946
Rohwer, S. (1977). Status signalling in Harris Sparrows: Some experiments in deception.
Behaviour, 61, 107–129.
Rohwer, S., & Rohwer, F.C. (1978). Status signalling in Harris Sparrows: Experimental
deceptions achieved. Animal Behaviour, 26, 1012–1022. doi:10.1016/0003-
3472(78)90090-8
Simmons, J., & Simonsohn, U. (May 2015). Power posing: Reassessing the evidence behind the
most popular TED talk. Retrieved from http://datacolada.org/2015/05/08/37-power-
posing-reassessing-the-evidence-behind-the-most-popular-ted-talk/
Simonsohn, U. (2015). Small telescopes: Detectability and the evaluation of replication results.
Psychological Science, 26, 559-569.
Stankowich, T., & Blumstein, D.T. (2005). Fear in animals: A meta-analysis and review of risk
assessment. Proceedings of the Royal Society B, 272, 2627-2634.
Tiedens, L.Z., & Fragale, A.R. (2003). Power moves: Complementarity in dominant and
submissive nonverbal behavior. Journal of Personality and Social Psychology, 84(3),
558–568. doi:10.1037/0022-3514.84.3.558
Wilson, A.D., & Golonka, S. (2013). Embodied cognition is not what you think it is. Frontiers in
Psychology, 4(February), 58. doi:10.3389/fpsyg.2013.00058
Running Head: POWER POSING AND DYADIC INTERACTIONS 27
Wilson, M.L., Britton, N.F., & Frank, N.R. (2002). Chimpanzees and the mathematics of battle.
Proceedings of the Royal Society of London B, 269, 1107-1112.
