THE BLOG
03/29/2016 05:33 pm ET | Updated Mar 29, 2016

The Reduction of Ego-Depletion

In recent years, psychological science has been increasingly roiled by a "replication crisis" wherein numerous theories the field was once certain about are proving rather wobbly on closer inspection. With the release of a new replication report by multiple scholarly labs in the journal Perspectives on Psychological Science, a theory called ego-depletion is the latest to come under scrutiny.

What is Ego-Depletion and Why Should You Care?

Ego-depletion refers to an idea that we have a limited amount of self-control or willpower. Every time we engage in one task that requires willpower, that results in less energy to do a subsequent task requiring willpower. For example if you convince yourself not to eat that piece of cake you've been eyeing, you may subsequently find it difficult to stay focused on a boring work task, or refrain from yelling at your partner when he or she annoys you. One classic experiment on the topic by Roy Baumeister and colleagues found that eating yucky food like radishes rather than chocolate or writing a speech expressing views that conflicted with one's personal attitudes tended to reduce subsequent persistence at a difficult puzzle task. People who had depleted their ego at one time period were less able to persist at a subsequent difficult task.

The basic ideas of ego-depletion have had remarkable penetration in the general public. Many life coaches and motivators include the concepts of ego-depletion in their work, there have been self-help books on the topic, and reports suggest that folks like President Obama and Facebook's Mark Zuckerberg wear basically the same colored clothes every day so that making a trivial choice on one matter won't reduce their ability to make more important decisions.

Note that ego-depletion, unlike some of the counter-intuitive or even silly sounding ideas from social psychology like Power Poses, has intuitive appeal. After all, who among us haven't noticed that, at the end of a long day, we no longer feel like making big decisions or dealing with fresh problems. However, this type of fatigue effect isn't quite the same thing as ego-depletion which posits depletion as drawing down on limited stores of will, not as a fatigue or mood related effect (although some scholars disagree with this distinction).

What Is Happening With Ego Depletion?

The trouble for ego-depletion began in earnest about a year ago with a meta-analysis (a research technique that combines multiple previous studies into a single analysis to gain a statistical overview of the research field as a whole). Although a previous meta-analysis had suggested that the ego depletion effect was robust, the new analysis suggested that results were mainly the product of publication bias (the tendency for journals to mainly publish papers supportive of a theory, and decline those that contradict it... an admittedly widespread problem in psychology and other fields). Correcting for publication bias suggested that the depletion effect was statistically no different from zero, although not all scholars agree with these methods of correction.

The new study is different. In this paper 23 different labs from around the world picked a single study documenting ego depletion. The study and protocol were agreed upon in advance both by the participating labs as well as the original study authors and Dr. Baumeister. Why replicate a single study so many times? Because with any one study, it's possible by random chance, one might get spurious results in either direction (either supporting or refuting a theory). With publication bias, only the successful chance findings get published which creates the false impression of a consistent field of results that may be due only to chance. Run, or replicate the same study multiple times, particularly with preregistration (in which all methods and analyses are published openly in advance, preventing researchers from monkeying with results after data is collected) and you have a pretty good sense of the "real" outcomes at least for a particular study. In this case, the ultimate results suggested that the "true" effect for this particular test of ego depletion is close to zero.

In commentaries on the replication effort, some concerns were noted. The original study authors, led by Chandra Sripada noted that some small differences between the original study and the replication efforts might have influenced results. That's a fair point, but, at best, suggests that ego-depletion is a fragile process, not one that can be easily applied to a wide variety of circumstances as it has been. Baumeister and colleague Katherine Vohs, commented that, in effect, they sorta agreed to the replication effort but not really. They state "Under the circumstances, we understood our approval to mean 'Sure, go ahead' and not 'yes, that's a definitive test of the phenomenon we've been studying all these years.'" They complain that several other procedures they had suggested were rejected by the replication labs. I'm less convinced by this commentary which feels a bit more after-the-fact sour grapes. Any concerns should have been clearly registered before the replication study, not after results were known. As they note, they thought that ego-depletion effects were robust enough that even the Sripada method should work... if it does not, even if one method doesn't represent all of ego-depletion, that's still notable.

So is Ego Depletion Done For?

Well, not exactly, but things don't look good. By itself, the replication study would only call into question the original Sripada study, not all of the research on ego-depletion. However, combined with the meta-analysis suggesting that ego-depletion writ large may have been more product of publication bias than true effects, then the theory is on the defensive. Baumeister and Vohs stated in their commentary that they're going to try to lead their own replication study. I'm a bit dubious how authentic such a study led by major proponents of the theory could be, but then again the current replication study in Perspectives on Psychological Science was hardly a hotbed of skeptical scholars, so perhaps I'm being too pessimistic. Preregistration of a Baumeister led replication effort will help prevent manipulating results once data is collected.

In the meantime it seems that ego-depletion is, at best, likely a shadow of the hyped version popularized in recent decades. Rather than a BIG THING that happens to everyone, ego-depletion may be a small thing that MIGHT happen to SOME people (but NOT many others), SOMETIMES, under the right conditions, but NOT others. The tendency of psychology to release dubious press releases with exaggerated claims is probably one of psychology's biggest failings as a field. We've been too brazen in taking tiny, transient, fragile trends and blowing them up into supposed groundbreaking finds that everyone needs to know about and incorporate into their own lives. The fall from those inflated heights can be embarrassing and sow reluctance to seriously consider new data that conflicts with a field's "sacred cows."

Psychologists will undoubtedly be debating ego-depletion for years to come. In the meantime, feel free to buy variously-colored shirts again. You'll be fine.

CONVERSATIONS