Contextual sensitivity in scientific reproducibility Jay J. Van Bavel\'. Peter M

Question

Contextual sensitivity in scientific reproducibility Jay J. Van Bavel'. Peter Mende-Siedlecki", William J. Brady", and Diego A. Reinerol of Prythology. New York Univeruity, New York, NY 1000 Tidited by Susan T. Fiske, Princeton Univernsity, Princeton, NI, and approved Agr75 2016 recelived for review December , 2019 n recent years, scientists have paid increasing atention to reprodus oncerms that these methodological insues in the fild ol peycholog chology, as well as in other fickk, is the roualt of a wide ange of aarprising ress (28, 29) In an efloet to enhance the of noscarch, several scicntific joumab (eg, Natue and hidden moderators") between the original research and the replication direct replications precsely these stualies use the same methods as those used in the original Many scientisds have also argued that the failure to repeoduce earch To help resolve this debate, we recoded the 100 original studies resalts might reflect contextual from the Reprodubky Poject on ??? extent to which the research moderators" between the original escatch and the replication topik of each study was contextually sensitive. Results suggested tht attempt (32-36) In fact, such suggestions peccede the current replication success, even after statistically adjusting for several method Touhey criticized a failed replication of his rescarch based on the tion between contextual sensitivity and replication success did not studied apart from the cultaral and hisdonical contexts that delme defer ??ss hologkal dpines. These results suggest that deir meanings (p S94 in ret 37) Indeed,the might that behavior earchen, replicator% and consum en should be mindful of antatua is a function d factors that might influence a paychological process We offer severall captured by Lewin's equation: B fUPE) (38) has shaped the have paid considerable attention to the influcnoe of context on the ndividual (eg, rets. 39-42) and have found extensive evidence that contcxtual factors alter human behavior (43-46) n recent years, scicntists have paid increasing attention to re- genetks (I) pharmacology (2), oncology (3, biology (4), and be a driving force behind scicntific inquiry and discovery. As nomics (5) have given credence to previous speculation that most statistician and political scientist Andrew Gelman recenily publishod research findings are fabse (6). Indeed, since the launch of the clinicaltrials gov registry in 2000, which forced researchers t0 Significance preregister their methods and outcome mcasares, the percentage of large heart-disease clinical trials reporting significant positive results plummeted from 57% to a mere 8% (7) The costs of such-Te- oter scientists. However, there is widespread debate in psy- producible preclinical research, estimatod at $28 billion in the chology (and other fields) about how to interpret failed repli United States (8), are staggering. In a sinmilar vein, psycholognt cations. Many have argued that contextual factors might have expressed growing concern regarding the reproducibilnty and account for several of these failed replications. We analyzed Scientific progress requires that findings can be reproduced by validity of paychological rescarch (eg, refs 9-14. This emphasis on 100 replication attempts in psychology and found that the reproducibility has produced a number of falures to repliate extent to which the research topic was likely to be contextually prominent studies, lcading professional societies and government sensitive (varying in time, culture, or location) was associated with replication success. This relationship remained a signifi cant predictor of replication success even after adjusting for such as the National Science Foundation to form agcncies subcommittees promoting more robust research practices (15) The Reproducibility Project in psychology mark in the scientific reproducibility movement. To help address viously had been associated with replication success (e.g ef has become a land characteristics of the original and replication studies that pre- in psychology, 270 researchers (Open fect size, statistical power). We offer recommendations for Science Collaboration. OSC) recenthy atlempted to directly repli- psychologists and other scientists interested in reproducibility cate 100 studies published in top psychology journals (16). Although the effect sizes in the original studics strongly predicted the effect Author ontributions 1L.VA, PM.S.Ws, and DAR sizes only 39% of pychology-dwayzed da -te- studies were unambiguously replicated (ie, were subjectively rated The uthons dectare no as having replicated the original result). These findings have been Ths wtde is PMAS Dwest Sabomiusion interpreted as a "bleak verdict" for the state of search (17). In turn, the results of the Reproducibality Project have o whom correpondence should be addresued tmail: jng wanbawilnyu.du deigned neeardh perf observed in replication atlempts teDuta depoition The data are available on the Open Sdience Framework the valuc of using psychology rescarch to This artectainssupporting information online at www.pnas.org/lookuphupplid ref. 18). This response corroborates recent tenmas?y711yOSupplementaL ome to question 64544459 | PNAS I kne7,2016 i vol. 113 I no.23

Explanation / Answer

Answer:

In the Introduction section of this publication, authors discussed the concerns regarding the irreproducibility of experimental results. Specifically to your answer, authors cited reference no 7, which reported the after forcing authors to preregister methods and outcome measures in clinicaltrials.gov registry in 2000, the percentage of large heart-disease clinical trials reporting significant positive results reduced drastically from 57% to a mere 8%. Which means the majority of positive reported results were actually negative. Authors further discuss that the burden of such non-reproducible preclinical research is about $28 billion in the United States alone.

Related Questions

Navigate

• Browse (All)
• Browse C
• Subjects

---

---

Integrity-first tutoring: explanations and feedback only — we do not complete graded work. Learn more.