Do psychology findings hold up? What a large replication study found, and what it means in Nepal

Researchers repeated 274 published findings from the social and behavioural sciences. Just over half held up. Here is what that means, and what research in Nepal adds.

You have probably seen headlines that begin with "Studies show...". A study says some habit makes you happier, or that one kind of person is more likely to do something. But how often does a finding hold up when other researchers test it again?

A large study published in Nature in 2026 tried to answer that question for the social and behavioural sciences. This post started from the American Psychological Association's Monitor summary of that study [2].

What is replication?

Replication means repeating a study, as closely as possible, to see whether the same result appears again. If a finding is real, a careful repeat should usually find it too. One study on its own can be misleading, by chance or because of how it was done. A finding that holds up again and again is easier to trust.

What did the study do?

Researchers from many labs tried to replicate 274 positive findings from 164 papers published between 2009 and 2018 in 54 journals [1]. The papers came from business, economics, education, political science, psychology and sociology [2]. The work was part of a United States research programme called SCORE (Systematizing Confidence in Open Research and Evidence) [2].

The repeats were designed carefully. On average they were large enough to detect the original effect, they used the original materials when these were available, and they were peer reviewed in advance [1].

What did they find?

The repeats found statistically significant results in the same direction as the original for 151 of the 274 findings, or 55.1% [1]. Counted by paper, rather than by finding, the figure was 49.3% [1]. Replication rates varied somewhat between fields, from 42.5% to 63.1% [1].

The effects were also smaller the second time. The typical original study reported a correlation of 0.25, while the typical repeat found 0.10 [1].

How you judge "success" matters too. The team compared 13 ways of deciding whether a replication had worked, and they gave estimates from 28.6% to 74.8% [1]. The researchers say there is no single, objectively correct way to decide whether a finding is replicable [2].

Does this mean psychology is broken?

No. The authors point out that some drop in effect size and significance is expected, partly because they only repeated positive results [1]. Studies that happened to find a strong effect by chance are more likely to look weaker when repeated.

The researchers say that understanding what affects how reliable research is can help decide when research findings are ready to be used in practice [2]. We read this as science checking itself, which is how it is meant to work. A finding that has been repeated is a stronger basis for advice than one that has not.

What does research in Nepal add?

Replication asks whether a finding holds up when it is tested again. In Nepal there is a second question: does a finding, or a questionnaire, work the same way here?

A test may not measure the same thing in Nepal. In a study of the Nepali version of a questionnaire about doctors' professional quality of life, 547 Nepali doctors took part [3]. The burnout part of the questionnaire did not hold together in this group [3]. The other two parts, on secondary traumatic stress and compassion satisfaction, were valid and reliable [3]. The authors' title says it plainly: burnout is largely a Western concept for Nepali doctors [3].

Help for children has not been tested well enough. A systematic review of child and adolescent mental health interventions in Nepal found nine studies, and their methods had weaknesses that affect how reliable they are [4]. Every study used some cultural adaptation, but the amount varied [4]. The authors call for high-quality studies that consider the collectivist nature of Nepali culture [4].

Nepali researchers are part of large shared studies. One way to test whether findings travel is to run the same study in many countries at once, with the plan registered in advance. Sandesh Dhakal and Narendra Singh Thagunna, of Tribhuvan University's Central Department of Psychology, are among the authors of a preregistered study across 58 countries with 18,128 participants [5]. In another study across 65 nations, a common life-satisfaction questionnaire largely measured the same thing in different countries and languages [6]. Sandesh Dhakal and Saphal Sapkota of KOSHISH are among its authors [6].

How can you read "studies show" more carefully?

These questions can help when you read about a psychology finding, in the news, on social media or in a class:

  • Is this one study, or many that agree?
  • How many people took part, and who were they?
  • Where was it done? Would it work the same way in Nepal?
  • Has anyone repeated it?
  • Did the researchers register their plan before collecting data?

None of this means research cannot be trusted. It means one study is a starting point, not the final word.

Sources

  1. Tyner AH, Abatayo AL, Daley M, et al. Investigating the replicability of the social and behavioural sciences. Nature 2026. https://doi.org/10.1038/s41586-025-10078-y (abstract read via Europe PMC)
  2. Mikulak A. In brief: Beliefs and intentions link self-control with healthy behaviors, and more scientific findings. APA Monitor on Psychology 2026;57(6) (item: "Replication rates across social and behavioral sciences hover near 55%"). https://www.apa.org/monitor/2026/09/beliefs-self-control-health
  3. Adhikari Y, Senft B, Poudyal B. Validity of ProQOL-5 Nepali version: burnout is largely a Western concept for Nepali doctors. NPRC Journal of Multidisciplinary Research 2025. https://doi.org/10.3126/nprcjmr.v2i8.83839 (abstract read via OpenAlex)
  4. Pacini A, Broker H, Shrestha P. Lost in translation? Cultural adaptation of child mental health interventions in Nepal: a systematic review. Journal of Child and Family Studies 2024. https://doi.org/10.1007/s10826-024-02880-5 (abstract read via OpenAlex)
  5. Kunst JR, Besta T, Jaśkiewicz M, et al. The psychology of offensive and defensive intergroup violence: preregistered insights from 58 countries. PNAS 2026. https://doi.org/10.1073/pnas.2535665123 (abstract and author list read via Europe PMC and OpenAlex)
  6. Swami V, Stieger S, Voracek M, et al. Life satisfaction around the world: measurement invariance of the Satisfaction With Life Scale (SWLS) across 65 nations, 40 languages, gender identities, and age groups. PLOS ONE 2025. https://doi.org/10.1371/journal.pone.0313107 (abstract and author list read via Europe PMC and OpenAlex)

Get news like this by email. Subscribe to PsyNepal updates →

Related reading

More from the PsyNepal blog

Find your way around psynepal.org

Also from PsyNepal: Master · Entrance · Self-checks · BMR · How they fit together

Need help now? Call 1166, free, 24 hours. More helplines