Tag Archive | replication study

“Fake It Till Make It” – Part II

In “Part I” I focused on “Adopting Power Posture to Feel Powerful”, and never thought I’d revisit the issue again. My previous post was largely based on Amy Cuddy’s research and TED talk in which she posits that a two-minute Wonder Woman’s power posture, hands on hips, would cause a spike in one’s adrenaline to accompany the assertion of confidence, which in turn leads to greater willingness to take risks in subsequent behavior. Many followers of her TED presentation have commented that the power posture really works and has helped them build their confidence. Her paper was published in 2010; her TED talk was in 2012, and her career subsequently soared in the areas of public speaking and consulting all over the world.

Since then, the field of social psychology – Cuddy’s professional base – has gone through a paradigm shift urging new guidelines for methodology that are much more rigorous. The champions of this shift, one of whom went to Princeton with Cuddy, challenged the field’s decades-long research practices for relying too much on convenience and tolerance of personal biases (thinking they are immaterial and wouldn’t affect research outcomes), and profoundly lacking in replication of studies. (Reinforcing this last point, journals have not been welcoming of “replication” articles.) These reformers suspected that most studies confirmed researchers’ hypotheses – false positive — and were insufficiently subjected to follow-up studies that would support or refute these confirmations.

This was a huge warning to the field of social psychology (and likely other branches of social sciences). The initial hostile reactions to the reformers’ challenge became louder and messy dynamics ensued. Dr. Cuddy’s work got ensnared in the back-and-forth. While there have been a few other prominent scholars whose studies have been questioned, none seemed to have suffered the same degree of browbeating and crashing of a promising career as Dr. Cuddy has. She has since withdrawn her tenure submission from the Harvard Business School.

The crux of the criticism of Dr. Cuddy’s initial work, the research on power pose, was: (1) the sample size of 42, 21 pairs of power pose and non-power pose, was too small, (2) the measurement of testosterone and cortisol levels was imprecise and/or inaccurate, (3) the danger of false positive results was not adequately addressed, and (4) at least one replication study done by a different researcher demonstrated that Dr. Cuddy’s power pose conclusion was weak at best.

My initial reaction to Dr. Cuddy’s power pose study was a sense of disquiet – for instance, can such a surge in confidence be sustained long enough to be beneficial? It was largely based on some philosophical considerations (see my previous post on the topic). The methodological issue is much more consequential; yet, the ensuing quarrels in the field have largely taken on philosophical, attitudinal, or social dimensions, and even touched on professional etiquette, far-removed from what a “simple” technical forum would entail. Personally, I think this reflects the messy nature of most social topics (consider the current storm regarding sexual assaults and harassments…or more mundane stuff like “performance evaluation”). I still think there is a lot of validity in the semi-jest term “physics envy” for describing social sciences.

A good portion of the ugliness in the criticism of Cuddy’s work came in the form of blogs and other social media. The three reformer-researcher-authors have their own blog, http://www.datacolada.org, that focuses on methodological and statistical issues, of which Cuddy’s work was one entry that has attracted readers across different disciplines. I understand and appreciate anyone who, in today’s internet-centric world, desiring more direct feedback and generating more discussion, takes it to blog format; however, there is a tendency for relaxed decorum in such a format. As one professor points out, “Because of social media and how it travels – you get pile-ons when the critique comes out, and 50 people share it in the view of thousands. That’s horrifying for anyone who’s critiqued, even if it’s legitimate.” (from New York Times Magazine article) Even one of the three authors for data colada, Joseph Simmons, who attended Princeton with Amy Cuddy, thought the treatment and criticism of Dr. Cuddy’s power-pose research has been unfair; after all, “the original study wasn’t particularly egregious. It was published in 2010 before anyone was thinking about this [the subsequent methodological revolution].”

It is difficult to objectively assess why Cuddy’s work was made to be the “poster child” for the criticism, especially without a rigorous study to survey professional reactions to Dr. Cuddy’s work and to compare the treatments of other top researchers’ pre-revolution studies. Even if we could do such a study, can we replicate it?! Sarcasm aside, I personally find the pile-on treatment particularly troubling. It is understandable that we all would defend our data, our methods, and our findings, for as long as possible; however, if enough subsequent evidence ultimately leads to a contradictory conclusion, no matter how uncomfortable for us personally, we learn to yield and move on. (This is the core of the scientific method and indeed the entire Scientific Revolution: We learn to accept the conclusions demanded by evidence.) Dr. Cuddy has grudgingly acknowledged her initial study’s flaws and tried to move on, so, why the personal attacks? Perhaps for social scientists themselves, it is not so easy to separate the work from personality, and for critics of social science, it’s much more fun to not even try?

While I was not a big fan of the power-pose study, it was intriguing. Now that I understand how weak the results are, I am fine with using the study as a springboard for further discussions. I simply do not understand why there is, and I emphatically deny any need for, adolescent rudeness and an infantile lack of decorum on social media in critiquing/attacking others’ work. Academics who would otherwise observe professional courtesy and standards in writing for journal publications become brutal, arrogant, and almost bullying on Facebook, Twitter, blog (personal or professional), or whatever form du jour.

It’s understandable that Cuddy’s research got the lion’s share of scrutiny – the bigger the reputation, the bigger the target – yet, it’s ironic that the very movement that tries to steer away personal bias seems driven by personal bias (why did they choose Cuddy’s work to focus on? Was it purely based on the hyperbolic reception of the work rather than the offending execution of the work?). It’s easier to target the most noticeable mistake, rather than the most egregious. So, these critics just threw their darts at an obvious target (TED talk has that effect, especially when it’s the second most popular one) rather than first making systematic assessment, say of other previous well-known studies. Isn’t that the very definition of selecting data based on convenience?

After finishing reading New York Times’ expose of this whole saga, all I can think of is: We sure like to scapegoat others for the very attribute for which we can be just as flawed. Shakespeare’s Hamlet scolds his mother for this: “Lay not that flattering unction to your soul, that not your trespass but my madness speaks.”

I used to feel apologetic for ending an article in a negative tone, and especially so close to our end-of-year holiday seasons. Times have changed…

 

Staying Sane and Charging Ahead.

Direct Contact: taso100@gmail.com

Advertisements