Ego depletion theory in trouble?

The field of psychology has been reeling because of problems with replicability, in that studies that claim to see certain effects have later had doubts cast upon them when efforts to replicate them failed to do so. Part of the problem is of course dealing with human subjects. But one of the theories that seemed to be pretty robust was based on a study by Roy Baumenister and Dianne Tice that suggested that people have a finite reservoir of will power and that when that is depleted by using it on some tasks, then we have reduced ability to overcome new challenges until that reservoir is replenished.

This article by Daniel Engber summarizes the original study.

Baumeister and Tice stacked their fresh-baked cookies on a plate, beside a bowl of red and white radishes, and brought in a parade of student volunteers. They told some of the students to hang out for a while unattended, eating only from the bowl of radishes, while another group ate only cookies. Afterward, each volunteer tried to solve a puzzle, one that was designed to be impossible to complete.

Baumeister and Tice timed the students in the puzzle task, to see how long it took them to give up. They found that the ones who’d eaten chocolate chip cookies kept working on the puzzle for 19 minutes, on average—about as long as people in a control condition who hadn’t snacked at all. The group of kids who noshed on radishes flubbed the puzzle test. They lasted just eight minutes before they quit in frustration.

The authors called this effect “ego depletion” and said it revealed a fundamental fact about the human mind: We all have a limited supply of willpower, and it decreases with overuse. Eating a radish when you’re surrounded by fresh-baked cookies represents an epic feat of self-denial, and one that really wears you out. Willpower, argued Baumeister and Tice, draws down mental energy—it’s a muscle that can be exercised to exhaustion.

Baumeister and others went on to suggest that the reservoir of will power was like a muscle and could be replenished and strengthened by time, sweet drinks, or mental exercises.

That paper spurred a lot of other studies that supported it and seemed to indicate that it was a robust effect applicable under a wide range of conditions. It also had the benefit of plausibility, a powerful factor in determining how easily people are persuaded by a new idea. We all know how tired we feel after we have been engaged in some task that requires the use of will power and how our resistance is reduced. When I was director of the teaching center, I used to advise faculty never to make an important decision just after finishing a lecture because it is at that time that one is mentally tired and thinking lazily, and can be tempted to opt for the easy fix that can cause problems later.

But Engber says that a researcher named Evan Carter looked more closely at all the studies and was not convinced that the effect was real.

A paper now in press, and due to publish next month in the journal Perspectives on Psychological Science, describes a massive effort to reproduce the main effect that underlies this work. Comprising more than 2,000 subjects tested at two-dozen different labs on several continents, the study found exactly nothing. A zero-effect for ego depletion: No sign that the human will works as it’s been described, or that these hundreds of studies amount to very much at all.

This is a big deal.

For scientists and science journalists, this back and forth is worrying. We’d like to think that a published study has more than even odds of being true. The new study of ego depletion has much higher stakes: Instead of warning us that any single piece of research might be unreliable, the new paper casts a shadow on a fully-formed research literature. Or, to put it another way: It takes aim not at the single paper but at the Big Idea.

It will be interesting to follow this issue. Baumeister is a very well-respected researcher and he and Tice did their original research while they were at my university. I got the chance to talk about this research in some detail when he returned a few years ago to give a talk on this. He is naturally is concerned that his major work is being called into question.

Baumeister, for his part, intends to launch his own replication effort, using methods that he thinks will work. “We try to do straight, honest work, and now we have to go to square one—just to make a point that was made 20 years ago. … It’s easier to publish stuff that tears something down than it is to build something up,” he told me wearily. “It’s not an enjoyable time. It’s not much fun.”

If it’s not much fun for the people whose life’s work has been called into question, neither does it hearten skeptics in the field. “I’m in a dark place,” Inzlicht wrote on his blog earlier this week. “I feel like the ground is moving from underneath me and I no longer know what is real and what is not.”

In researching my forthcoming book on the logic of science, I will be using this episode as an example of a common phenomenon in science, of a theory taking deep root in a field and then being questioned. How its supporters react to the challenge, and how a new consensus emerges is a fascinating process.


  1. Cuttlefish says

    How many papers were published on N-Rays?

    I, for one, am very happy to question the ego-depletion stuff. The “plausibility” always seemed more folk-psychology than science, so I will be watching to see how this all shakes out.

  2. Robert B. says

    Were they spicy garden radishes like my mom used to grow, or boring tasteless supermarket radishes?

    Also, giving up on an impossible puzzle is a success, not a failure, and it seems like the original experiment was based on the assumption that the subjects didn’t know that. But I can totally see kid-me trying a puzzle a couple times, then taking five minutes to prove it impossible, and then quitting. And quite a lot of kids I work with now would be able to guess that the stupid adults had handed them an impossible task. Maybe being told to eat radishes while everyone else gets cookies erodes one’s trust in authority.

    I like the theory of ego depletion, it seems to explain a lot about my experiences and symptoms, but on reflection I don’t like that study much.

  3. flex says

    I’ve never really been that fond of the ego depletion theory as presented. In the original study there are confounding variables like “we were rewarded first for this task! let’s keep going!” and “Sugar High!”, which didn’t really feel like they were controlled for. Then it is a theory about a mental aspect which is largely ill-defined (how do you define will-power?) and doesn’t correlate well to other known physiological aspect of the brain even though the theory predicts a typical physiological response.

    But my biggest problem was that it never aligned to my own experiences.

    I think we all have experienced the feeling of being mentally tired, depleted, and unable to make decisions.

    However, my own experiences suggest that it’s not the act of exercising willpower which make my mind want to always take the easy solution, but the amount of resistance I get to the decisions I make (or have made) which makes it harder and harder to make the next decision.

    After 20+ years as a product engineer in a tier 1 company in the automotive industry, I find that if I make a decision which everyone thinks is good and acceptable, it’s easier to make the next one. Even if that original decision required a lot of mental activity.

    However, when I come to a conclusion that no one wants to hear, I then get a great deal of resistance in accepting the decision I made. I have to explain why I came to that conclusion. First to my team, then to my management, then to my customer, and often then to my customer’s management. And every time there are people challenging my conclusions and regularly there are a few who simply reject those conclusions.

    That is what makes go home and be unable to put the books back on the shelves in my library.
    That is what makes me unable to cook a meal when I leave work.
    That is what makes me want to have a martini and sit in a chair staring off into space.

    It’s not the decisions which get to me, it’s the continued questioning of those decisions, sometimes for months and years, after they are made. That saps my willpower. Once the workday is over I have little desire to even untie my shoes. I certainly don’t have the energy to do something which will release me from this grinding job.

    They do pay me well though, so I guess, even though everything I do is questioned, that I’m doing a good job.

  4. Holms says

    I’m amazed the thesis was considered plausible in the first place, let alone becoming enstablished as a ‘Big Idea’. It seems to me that the study ‘found’ that serially screwing with people leads to those people becoming frustrated, i.e. exactly what we would expect when we go out of our way to annoy someone. It seems the central idea amounts to ‘be annoying ==> people get annoyed’. Whoopdy do.

  5. says

    Psych experiments also suffer hugely from sampling bias, especially the ones studying college students as if college students represent all of humanity.

    ‘Ego Depletion’ is based on an unmeasurable ineffable ‘ego’ thing that in turn is based on a bunch of stuff Freud threw out his asshole on a fine day. The endgame on that ought to have been obvious from day 1. When you ask a psychologist “what is ego?” they’ll eventually loop back into a circular definition (the clever ones, who realize what you’re doing, will lead you a merry chase) Eh. Skinner tried to focus psychology on observables, and to get away from hypothesizing mechanisms with no plausible behind them, but even then psychology tends to generalize rather stupidly: Oh, if rats behave in this way maybe we can infer how humans will behave in a similar circumstance. Nonsense.

    Psychology is eventually going to get absorbed into neuroscience, except for a few bloviators who become full-time woo peddlers and who aren’t capable of understanding neuroscience or who are too unbearable to work with.

  6. Nick Gotts says

    It’s easier to publish stuff that tears something down than it is to build something up -- Baumeister

    This strikes me as the precise reverse of the truth: getting failed replication studies published is notoriously difficult.

  7. says

    It seems the central idea amounts to ‘be annoying ==> people get annoyed’. Whoopdy do.

    Perfect example of the kind of circularity I was referring to in my other comment. What is ego? It’s the thing that controls you rationally! What is rational control? It’s your ego, of course!

    The “be annoying --> people get annoyed” isn’t even good science (I realize you’re parodying) because it’s going to be hugely culture-dependent. I would trivially argue that ‘culture’ is one thing humans do to manage annoyance (among many other things) so we should expect that a !Kung or a Yanomamo would respond to that experiment differently from a Trump supporter or a buddhist or a starving undergraduate who’s just trying to get the $50 for the experiment so they can buy some weed.

    Psychology really needs to be uprooted and plowed with salt; these results are the beginning of that process. It’s going to get worse for psychology before it gets better (and that will look like: buhbye psychology!) When I was a psych undergrad they were still talking about freudian theory, FFS. And I don’t mean in a historical context, like how physicists talk about N-rays (“ooh, we learned not to fuck up like that..”) There’s waaaay too much cataloguing going on, which results in circular definition of stuff. What is a psychopath? It’s when you demonstrate some of the following characteristics. Oh, so it’s an inventory and you have absolutely no cause/effect relationship you can establish? We’re done.

    This reproducibility attack is the beginning of the end for the field. The end will be that things like DSM will be overthrown as well -- which is a big problem in lots of ways because health insurance funding occasionally depends on it. But stuff like DSM is the catalog of characteristics — and it changes because of prevailing politics. Actually having a disorder doesn’t. I don’t think psychologists should even be allowed to diagnose “disorder”s because a disorder implies an underlying condition in a cause/effect relationship. Simply cataloging behaviors doesn’t let you accurately infer a disorder is present unless a certain behavior always indicates the disorder … The problem is that the psychologists are comfortable prescribing massive amounts of psychoactive pharmaceuticals based on their catalogues of behaviors. That whole edifice ought to come sliding down but there’s a solid chance that neuroscience will come along and shore it up before it completely collapses. At which point psychology will declare victory.

    If you want to torture an intro psych professor, when they start talking about Maslow’s hierarchy of needs, ask them how “self-actualization” is qualitatively different from “happiness” and then ask them whether they’re saying “if you have all the shit you need, you’ll be happy”?

  8. says

    PS to my above, re: DSM
    All that’s necessary is a good study showing that DSM does not produce predictable diagnoses across a population. Conclusion: psychology doesn’t know what it purports to be measuring. Game over. Nuke it from orbit.

  9. lanir says

    Caveat: I didn’t read the linked article, seemed lengthy and I can never focus on psychology very well without dreaming up my own ideas.

    I think this would be tested more reliably and with less variables by having the same people come in once a day for a set period of time to do some repetitive task but one that requires enough infrequent thought that they can’t zone out for it easily. Essentially this is testing one’s willpower to do makework that has no real purpose and measuring the degree of performance one manages to pull off.

    Each day they have an envelope to open before they begin. Inside the envelope are various things that are designed to add more or less stress. Perhaps one day it’s a blank paper. Another it’s a reminder that the weather is great outside but the subject can’t leave early. The last day maybe it’s the payment for being in the study or a reminder that they’ll be paid at the end of the day or something. Control group could either get no envelope or all blank pages; or do both and see if there’s a difference.

    Compare each person’s day to day results using either time or amount completed, whichever the study authors prefer. See if and how their progress changes versus the control group(s) average.

    I think it would have the advantage over the radish/cookie experiment in that it would remove the differences in how people resolve puzzles. The disadvantage is you’d have to do some hard thinking about the contents of the envelopes to make them broadly meaningful to different subjects. Also it would take time and not provide cookies to consume at the end. These are serious disadvantages to consider.

  10. Trickster Goddess says

    I first had to read some psych experiments for a tangentially related course in university and right from the start they drove me nuts at how unscientifically they were conducted. Premises would be based on assumptions, there would be 2 or 3 controlled variables, but would overlook a dozen other uncontrolled one, and it would be all boiled to a couple of narrow conclusions. Oh yeah, and it usually involves lying to the participants.

  11. Trickster Goddess says

    I thought this footnote on the original paper was quite laughable:

    2 As this article went to press, we were notified that this experiment had been independently replicated by Timothy J. Howe, of Cole Junior High School in East Greenwich, Rhode Island, for his science fair project. His results conformed almost exactly to ours […] These converging results strengthen confidence in the present findings.

    Yes, nothing boosts the credibility of your results like having them backed up by a junior high student.

  12. johnson catman says

    So, are the egos of Baumeister and Tice being depleted by the studies contradicting their study?

  13. AbstractIndescretionist says

    Good riddance to the American Psychological Association who supported the Rumsfeld/Cheney torture regime. And may the names and reputations of war-criminals/psychologists James Mitchell and Bruce Jessen live in infamy. The sooner the discipline/study of psychology is forgotten the better.

  14. anat says

    So what becomes of the advice we were given that was based on ego depletion? Is it in fact not a bad idea to shop for groceries while hungry?

Leave a Reply

Your email address will not be published. Required fields are marked *