Case Study 1: The Ego Depletion Saga — The Rise and Fall of a Textbook Finding

The Claim

In 1998, Roy Baumeister and colleagues published a study claiming that willpower operates like a muscle: it fatigues with use. Participants who resisted eating freshly baked cookies (an act of self-control) performed worse on a subsequent puzzle task. The interpretation: self-control draws on a limited resource that gets "depleted" — hence "ego depletion."

The Rise

The ego depletion finding was a sensation. It was intuitive (everyone has experienced feeling drained after exercising self-control), experimentally elegant (the cookie/radish paradigm was simple and replicable-seeming), and practically relevant (it had implications for dieting, addiction, decision-making, and productivity).

Over the next 15 years: - Over 200 studies were published on ego depletion - The concept appeared in virtually every social psychology textbook - Popular books (Willpower by Baumeister and Tierney, 2011) brought the idea to millions of readers - Corporate wellness programs incorporated ego depletion into their advice - The finding accumulated thousands of citations

The Fall

In 2016, a massive multi-lab replication effort — involving 23 laboratories and over 2,000 participants — tested the core ego depletion effect using a pre-registered protocol. The result: no evidence for ego depletion. The effect size was essentially zero.

Subsequent meta-analyses, incorporating publication bias corrections, reached similar conclusions: the published literature dramatically overstated the effect. When corrected for publication bias and p-hacking, the ego depletion effect shrank to near zero.

What Went Wrong

The ego depletion case is a textbook illustration of every mechanism discussed in this chapter:

  • P-hacking: Researchers tested multiple outcome measures and reported only significant ones
  • Publication bias: Studies finding no ego depletion were filed away; only positive results were published
  • Small samples: Many ego depletion studies had 20-40 participants per condition — insufficient to detect small effects reliably
  • Researcher degrees of freedom: The "ego depletion" manipulation and the outcome measures varied enormously across studies, allowing researchers to find "depletion effects" in many different forms
  • Narrative appeal: The metaphor of willpower as a muscle was so compelling (plausible story problem, Ch.6) that researchers and reviewers were predisposed to believe positive results

The Sunk Cost Challenge

When the multi-lab replication was published, Baumeister and colleagues initially resisted the findings, arguing that the specific depletion task used in the replication was not the best one, that the effect required specific conditions to appear, and that the replication protocol was too rigid. These objections may have some merit — but they also illustrate the sunk cost dynamic: when your career's most cited finding is at stake, the bar for accepting disconfirmation rises dramatically.

Discussion Questions

  1. The ego depletion concept was intuitive and practically useful even if the specific experimental evidence was unreliable. Does the replication failure mean the concept is wrong, or just that the specific studies were unreliable?
  2. Over 200 studies were published supporting ego depletion. How could an entire literature be wrong? What does this tell us about the self-correcting nature of science?
  3. Apply the "Would this replicate?" test retrospectively to the original cookie/radish study. What features should have raised concerns?
  4. If you had built your career on ego depletion research, how would you respond to the replication failure? What does the chapter suggest about the structural forces you would face?

References

  • Baumeister, R. F. et al. (1998). "Ego depletion: Is the active self a limited resource?" JPSP, 74(5), 1252–1265. (Tier 1)
  • Hagger, M. S. et al. (2016). "A Multilab Preregistered Replication of the Ego-Depletion Effect." Perspectives on Psychological Science, 11(4), 546–573. (Tier 1)
  • Carter, E. C. et al. (2015). "A series of meta-analytic tests of the depletion effect." JPSP, 109(4), 703–723. (Tier 1 — meta-analysis showing publication bias inflation)