The Hidden Costs of Academic Perfectionism
When "just one more study" goes too far.
Posted May 12, 2025 | Reviewed by Gary Drevitch
While my posts typically focus on social isolation and loneliness, today I'm discussing something equally important: the often-hidden costs of pursuing scientific knowledge.
As scientists, we strive for rigor. We want to be thorough, careful, and precise. But at what point does thoroughness become excessive? When does the quest for scientific perfection become wasteful—of time, money, and human potential? How do we know when enough is enough? A recent replication project I supported when I still worked at the university offers a sobering case study of these hidden costs.
In 2018, Kevin Vezirian, who had just finished his master's degree, began what should have been a straightforward replication of a 2013 article published in the Journal of Experimental Social Psychology . The original study claimed that seeing black-and-white patterns makes people's moral judgments more extreme—a visual "black-and-white thinking" effect that the authors framed as universal. Kevin asked me to support this project, and what began as a simple investigation quickly became a saga that would span six years, consume thousands of work hours, and cost over €20,000.
The original study had serious methodological limitations from the start. With only 111 participants divided across three conditions, it was statistically underpowered to detect the effect size it reported. Despite these limitations, the authors had confidently claimed a universal psychological effect.
The evidence against the original effect was already compelling by July 2018. Kevin's initial paper reported three pre-registered replications (total N = 846) in which only one of the three studies showed an effect similar to the original. A meta-analysis across all three replications and even including the original study showed no significant overall effect.
But the scientific publication process demanded more. After the initial three studies, we conducted a large-scale study with French participants (N=8,602)—a sample 77 times larger than the original. Our Bayesian analyses showed that the absence of the effect was 33 times more likely than its presence.
Yet the demands continued. Journal reviewers insisted on more evidence—specifically, a replication with English speakers to match the original UK sample. Despite complying and finding the same null results with English participants (N=365), reviewers continued to request additional analyses and experiments.
The final paper was only accepted in late 2024—nearly six years after the project began and the initial evidence was available.
Our team of four collaborators spent a total of 1,100-1,500 hours on this replication. Converting these hours to standard French academic salary rates yields approximately €19,000 in labor costs alone—and this is based on French academia, where researchers are significantly underpaid compared to many other European countries, with foreign researchers often receiving even lower compensation than their French peers. Had this study been conducted in the United States, the same labor hours would have cost $45,000-60,000, given that American assistant professors earn roughly 2-3 times what their French counterparts do.
Add to this the direct expenses: €204 (£169) for participant payments through Prolific, approximately €1,000 for our share of institutional software licenses, and various administrative costs. The total: approximately €20,200.
What makes this cost particularly troubling is what it bought: evidence that a theoretically vague and practically negligible effect likely never existed in the first place. The original "black-and-white thinking" effect, even if it had been real, would have had minimal implications for real-world behavior.
Most frustratingly, over €16,000 of these costs were incurred after July 2018, when the evidence against the effect was already statistically compelling. Had the scientific community been willing to accept strong pre-registered multi-study evidence earlier, these resources could have been directed toward more promising research questions.
Impact on Early-Career Scientists
This burden falls especially hard on early- career researchers. For Kevin, the project began as an educational opportunity to learn about replication science. What should have been a stepping stone into rigorous psychological science became a years-long slog through academic bureaucracy.
To be clear, this isn't an argument against rigorous science; quite the opposite. Kevin tried to do everything right: His studies were preregistered, had more than twice the statistical power of the original study, employed proper statistical analyses, and maintained full transparency with materials and data.
This case exemplifies a deeper problem in our scientific culture: We often demand disproportionate evidence for null findings compared to positive ones.
The original study, despite its statistical fragility and methodological limitations, was published in a high-impact journal based on two studies with barely significant results and wide confidence intervals. Yet to establish that this tenuous effect likely doesn't exist, we needed multiple high-powered studies, sophisticated Bayesian analyses, and cross-cultural evidence—all at substantial cost. While reviewers' requests for additional data had specific rationales—addressing relatively minor points like the battery of questions preceding the main task and "moral-related items" in the battery—the cumulative burden of these requests far outweighed their scientific value.
I'm not suggesting we lower our standards. Rather, I propose proportional scientific scrutiny that considers three factors:
The standard of evidence required to disprove a fragile effect should be proportional to the strength of the evidence that established it.
Based on this experience, I see several practical changes that could help:
Science isn't just about p-values, effect sizes, and methodology. It's a human endeavor carried out by researchers with limited time, resources, and energy. When we forget this, we risk creating an unsustainable system that drains rather than nurtures talent.
Kevin's six-year journey to publish a "simple" replication revealed not just the absence of one specific psychological effect, but the deeper dysfunction of a scientific culture in which doing good work is not enough.
Perhaps that's the more important finding.
Dr Kevin Vezirian is now a Maître de Conférences (Assistant Professor) at Université Savoie Mont Blanc (LIP/PC2S) in Chambéry, France.
Share this post Facebook Bluesky Linkedin Email
There was a problem adding your email address. Please try again.
By submitting your information you agree to the Psychology Today Terms & Conditions and Privacy Policy
Hans Rocha IJzerman, Ph.D., is Founder and Director of the Annecy Behavioral Science Lab (ABSL) in France and is a Research Affiliate at the University of Oxford.
Get the help you need from a therapist near you–a FREE service from Psychology Today.
This article is part of the Bringwise Psychology Journal — daily insights on human behavior, mental health, and personal growth.