Here Is Amy Cuddy’s Response to Critiques of Her Power-Posing Research

Photo: Craig Barritt/Getty Images for WME Live

As Science of Us reported earlier this week, last Sunday night the psychologist Dana Carney, co-author of the 2010 study that gave rise to the concept of “power posing” — that is, adopting and holding certain open and assertive postures as a technique for improving performance in business and other settings — posted a surprising note on her faculty website in which she stated that she didn’t think that the power-posing effect was real, and that the original study had fatal methodological shortcomings. In that study, participants who power-posed had increases in testosterone, decreases in cortisol (the stress hormone), and increased risk-taking in a gambling task, as compared to members of a group who adopted meek postures. According to Carney, those findings were likely the result of so-called p-hacking and other statistical shortcuts taken by the research team (she argued that, at the time, those practices were less frowned-upon than they are today).

Within the psychological community, the note was seen as potentially damaging to Amy Cuddy, the Harvard Business School psychologist and co-author on the original study who has built a very successful career, including a blockbuster TED Talk, out of advocating for power posing and similar techniques. At 2:30 this afternoon, a publisher at Cuddy’s publishing house, Hachette Book Group, sent out a statement on Cuddy’s behalf to Melissa, and presumably to a number of other journalists as well. It marks Cuddy’s first public response to the Carney document, and it is presented below in its entirety, with all formatting intact.


From Amy Cuddy:

I was surprised by a recent statement that the power pose effect is “not real” and I want to set the record straight on where the science stands.

First, it’s necessary to make a critical clarification: There are scores of studies examining feedback effects of adopting expansive posture (colloquially known as “power posing”) on various outcomes. But I’m seeing the term “power posing effect” being tossed around to refer to many different phenomena. The key finding, the one that I would call “the power posing effect,” is simple: adopting expansive postures causes people to feel more powerful. Since my coauthors and I first published our evidence,this effect has been replicated in at least nine published studies and in at least four unpublished studies from nine different labs. In fact, the “power posing effect” – that adopting expansive postures causes people to feel more powerful – was even replicated in the pre-registered Ranehill et al. study with 200 participants in Switzerland, which strangely has been described as the “non-replication” of power posing. The other outcomes (behavior, physiology, etc.) are secondary to the key effect.

My lab recently conducted a systematic review and statistical analysis of power posing studies. It shows strong and robust evidence that adopting expansive postures does indeed increase feelings of power (manuscript under review). In addition, many studies have shown that adopting expansive postures increases happiness, mood, and other related emotion variables. We did not include those studies in the analysis of the main power posing effect.

I have confidence in the effects of expansive postures on people’s feelings of power — and that feeling powerful is a critical psychological variable. As Columbia University Professor Adam Galinsky and colleagues wrote in their 2016 review, a person’s “sense of power…produces a range of cognitive, behavioral, and physiological consequences,” including improved executive functioning, general optimism, creativity, authenticity, the ability to self-regulate, and performance in various domains, to name a handful.

The current complete body of evidence on expansive postural feedback, to my knowledge and based on a systematic review, includes 46 studies from 96 researchers from university labs around the world, representing various subdisciplines of psychology, including social psychology, health psychology, clinical psychology, and sports psychology. These studies look at the effects of adopting expansive posture on job interview performance, food intake of people with symptoms of disordered eating, hormones, motivation, learning, and leadership behavior, among many other outcomes. Again, our statistical analysis of these 46 studies reveals strong support for the effects of power posing on various psychological, behavioral, and physiological outcomes (manuscript under review). Note that I am not including the scores of studies on the psychological, behavioral, and physiological effects of other postural manipulations, such as upright versus supine posture, which has been shown to affect cortisol levels (Henning et al., 2000), potassium levels (e.g., Shirreffs and Maughan, 1994), calcium levels (Husdan et al., 1973), and heart rate variability (Vuksanovic et al., 2005), among many other physiological outcomes; nor did we include the hundreds of studies showing the psychological and physiological feedback effects of breathing, movement, yoga, and so on. As acclaimed neuroscientist Peter Strick recently said about his new PNAS study that identified the cortical mechanisms involved in core muscle engagement (the muscles we use to hold expansive postures) and stress reduction, “How we move, think, and feel have an impact on the stress response through real neural connections.”

I’m also interested to see two other very recent exploratory studies that have examined the effects of adopting expansive postures in populations that tend to lack power. In a study on the effects of power posing among older people, conducted by Pei-Lee Teh and colleagues, power posing increased older people’s willingness to learn about and use technology that can be helpful to them. In another study that is currently in press, conducted by Elizabeth Broadbent and colleagues, upright posture influenced the emotional states of people suffering from depressive symptoms, improving their mood and decreasing their fatigue.

Science grows incrementally, moved forward by improved data and discussion, often involving disagreements among scientists, which is a critical part of how science advances. Cancer treatment, for example, has expanded from chemotherapy to immunotherapy, based on broader scientific understanding, yet not all scientists and cancer specialists would agree on the best treatment in a given case, which is why people are encouraged to seek second opinions. So I welcome constructive back-and-forth examinations of research findings as we seek the truths that benefit society as a whole.

In this spirit, during the last academic year, I hired a statistician and post doc at the Harvard Institute for Quantitative Social Science to conduct an independent audit of our analyses in our 2010 article (Carney, Cuddy, & Yap, 2010), with the aim of thoroughly reviewing and documenting any and all potential errors. For complete transparency, I asked him to post his conclusions in an open forum without my involvement or review prior to publication. His summary, “Replication Data for “Power Posing: Brief Nonverbal Displays Affect Neuroendocrine Levels and Risk Tolerance” by Carney, Cuddy, Yap (2010),” has been live on the Harvard Dataverse since May 2016.

As a result of this independent audit and additional peer feedback, all the paper’s coauthors did agree that a correction was in order and together we were collaboratively and currently preparing a corrigendum (a published correction) of our 2010 Psychological Science article. (For those interested in this level of detail, the corrections concerned a small mistake in the identification of statistical outliers on testosterone and cortisol, which does not change the findings, and the p-value for one of the two reported risk-taking measures, which is .049 for the likelihood ratio test versus .052 for the chi-square test.)

I also cannot contest the first author’s recollections of how the data were collected and analyzed, as she led both. By today’s improved methodological standards, the studies in that paper — which was peer-reviewed — were “underpowered,” meaning that they should have included more participants. I wish we had conducted those studies with the rigor of today’s methodological standards, which I firmly believe are moving our field in the right direction. However, in no way does it invalidate the conclusions of nearly fifty studies on the feedback effects of adopting expansive postures. Even when excluding the 2010 paper, my lab’s systematic statistical analysis of the current body of literature supports the existence and robustness of power posing effects.

Like all scientists, I understand that my field evolves as new evidence replicates some effects and not others. The key power posing effect has been replicated at least nine times, even in Ranehill and colleagues’ well-powered study with 200 participants. Yet the Ranehill study did not replicate the hormone shifts and risk-taking effects. My scientific evaluation of the various effects will update as new information comes in.

When I speak about this work now, I talk about new evidence, replications, and non-replications. I highlight which effects I believe to be strongest and most important, and those about which we don’t yet know enough. For example, while I am confident about the key power posing effect on feelings of power and the overall evidential value of the literature, I am agnostic about the effects of expansive posture on hormones. The jury is still out. We have conflicting evidence, which is fascinating and means it could go either way. I’m intrigued to see where the new evidence will lead as research continues in social psychology and neuroscience.

As a person who has studied the burdens of marginalized groups and individuals who lack power, my research has centered on stereotyping, prejudice, and the discovery of psychological tools that can help people to feel more powerful. I am continuing this work. While I’m excited about the ongoing work on postural feedback effects, movement, breathing, and other body-mind interventions that have the potential to help people live richer, more rewarding and actualized lives, I’m also engaged in my new lab- and field-based experimental work on the emergence and importance of gender differences in the posture of children (why do our daughters begin to adopt collapsed behavior around middle school, and how does this affect them?), my continuing work on how adopting expansive versus contractive postures affects how others judge us, my research on how culture affects the contents of gender stereotypes, and developing research-based workplace interventions that improve the well-being and health of low-wage workers.

Finally, I am concerned that the tenor of discussions like the one that has been unfolding on power posing, and the tendency to discount an entire area of research on the basis of necessary corrections or differences between scientists’ assessments, may have a chilling effect on science. In the last six months, three labs have contacted me to let me know that they have conducted high-powered studies on expansive postures or “power posing,” but that they are reluctant to submit them because they fear being “targeted” for doing research in this area. Open science must be inclusive. As Atul Gawande put it, “The scientist has an experimental mind, not a litigious one.” We must consider the new and the old. The replication successes with the replication failures. This is how science evolves and knowledge emerges.

I, and all scientists and consumers of science, must reserve the right to update and revise their understanding and evaluation of phenomena as more evidence becomes available. Surely, each of the many researchers who studies the effects of posture has his or her own consideration of the body of research, how strong the effects are, what’s driving them, what are their boundary conditions, etc. This continuing study and analysis is how science moves forward, and I can’t imagine my own assessment will remain static. With new information, it will be updated.