March 8, 2016

The curtain falls for ego depletion

Last year I wrote that the existence of ego depletion is doubtful (here) and that I think the concept is based on a too simple way of thinking about psychology (here - sorry it is in Dutch). Now it appears that the curtain definitely falls for the concept of ego depletion.

In an article in Slate Daniel Enger describes how the concept has lost its credibility in a few steps. The article describes how a replication has been done of a study which showed the ego depletion effect (this study). As an aside I'd like to mention that the people who did the replication can hardly be suspected of trying to deny the existence of ego depletion. Martin Hagger led the replication study. In 2010 he did a meta-analysis which demonstrated the existence of ego depletion. By the way, this result was contradicted in a 2015 meta-analysis which was done more rigorously in several aspects. Roy Baumeister, the big man behind the ego depletion concept was also involved in the replication study. He was consulted with respect to methodological matters.

The study which was replicated used a self control task. Respondents looked at a screen on which simple words were briefly shown. They had to press a button when the word which was displayed contained the letter 'e' but only when is was not followed within positions by another vowel. The original study demonstrated a strong ego depletion effect. The respondents performed much worse on a second task.

In 24 labs around the world the replication study was done. The result: in 21 of the studies no effect was found, in 2 an effect was found, in 1 a reverse effect was found. Upon learning about these results, Roy Baumeister is not convinced that this shows that ego depletion does not exist or is not very important. He thinks that the way in which the study is done is the reason the effect was not found and is preparing his own replication study. Another psychologist who has done much research into ego depletion, Michael Inzlicht, from the University of Toronto, responds quite differently.

In a recent blog post he sounds down-hearted and he writes: 
"I have spent nearly a decade working on the concept of ego depletion, including work that is critical of the model used to explain the phenomenon. I have been rewarded for this work, and I am convinced that the main reason I get any invitations to speak at colloquia and brown-bags these days is because of this work. The problem is that ego depletion might not even be a thing. By now, many people are aware that a massive replication attempt of the basic ego depletion effect involving over 2,000 participants found nothing, nada, zip. Only three of the 24 participating labs found a significant effect, but even then, one of these found a significant result in the wrong direction!"


  1. In a replication study should all the manipulations and conditions of the original study be replicated? As I read in the cited article, it was not the case at least in some of the replications (use of computers instead of handwriting; not making mind habit first to be destroyed afterwards etc. etc. For example, it is difficult for me to imagine that in all the labs people bake their own cookies that disperse the heady aroma of the freshly baked and distract the subjects who only are allowed to eat radishes. Also, I think it depends where exactly do you do your experiment. In some countries the self-control the everyday behavior is very strong, for example in the countries in Northern Europe. This is not the case in southern countries or, say, India. Well I do not say that the theory of ego-depletion is correct - I only say "Do we really replicate when we replicate?"
    If we try to replicate Freud studies and observations, it would be very difficult or impossible, I think, because the society was changed by his theories.
    Anyway, the problems of replication in non-sciences (a.k.a. social sciences in order not to be mistaken with the real sciences) are insignificant compared to the problems of physicists: it seems the proton shrank in size

    1. I think your general question regarding replications is important. When do we speak of a real replication?

      By the way, I do not agree with your distinction of real sciences and non-science. I know that some people, like Feynman, have suggested that the social sciences are not real sciences. I don't think this is a valid distinction. Good social science is the application of the scientific approach to social matters. I do not see what is non-scientific about that. Saying that social science is non-science is to pessimistically declare that science cannot be applied to this domain. I have never heard a convincing argument why this should be so.


Enter your email address:

Delivered by FeedBurner