Publication Decisions and their Possible Effects on Inferences Drawn from Tests of Significance—or Vice Versa

“After all, Sterling (1959) found zero replications in the total of 362 articles piiblished col- 676 O. w SMITH, ET AL. lectively in 1955 by the Journal of Experimental P~ychology, Comparative and Physiological Psychology, J o z ~ r m l of Clinical P~ychology, and Jozlrnal of Social Psychology.”

“While this mechanism of suppressing non-significant results, called the file drawer effect, and its impact on the published literature, has been identified long ago in theoretical models (e.g., Ioannidis, 2005; Rosenthal, 1979; Rozeboom, 1960; Sterling, 1959), it has been ignored by economists for a long time, especia…”

“Например, A.G. Greenwald утверждает, что значимые/достоверные (с точки зрения статистических сдвигов) результаты будут в восемь раз вероятнее опубликованы, чем незначимые/ недостоверные (Greenwald, 1975), и 97% статей в научных периодических психологических журналах подтверждают это утверждение…”

“File drawer bias A file drawer bias is based on the idea that researchers more frequently publish experimental findings that reject the null hypothesis (Rosenthal, 1984; Scargle, 2000; Sterling, 1959).”

“Furthermore, the practice of publishing only significant results causes publication bias (Sterling 1959).”

“…in psychology, including journal editors and referees, tend to confuse nominal statistical significance of an experimental finding with its epistemic import (e.g., Rosenthal & Gaito 1963; Neuliep & Crandall 1990, 1993; Sterling 1959; Sterling, Rosenbaum, & Weinkam 1995; Gigerenzer 2004, 2018).”

“The consequences of the file drawer problem for the published evidence base were recognized decades ago (Sterling, 1959 ; Greenwald, 1975 ; Rosenthal, 1979 ; Hedges, 1984 ), yet it remains a persistent issue despite sustained methodological debate and reform efforts.”

“…resonates with the argument that selective visibility where only superior or favourable results are published distorts the methodological archive, a concern raised by Ioannidis [4], Sterling [5], and Loken and Gelman [6] in their analyses of publication bias and noise-driven inflation of effects.”

“When such results are selectively reported, the resulting inflation of t -statistics or related test statistics reflects p -hacking (Simmons et al., 2011) and contributes to publication bias (Sterling, 1959).”

“No particular mechanism”

“; Rosenthal (1979); Sorić (1989); Sterling (1959) argued that bias occurring from publication policies that favor statistical significance, called publication bias, could lead to a high false discovery rate (FDR) among published results.”

“An inadequate or incorrect introduction of students to the scientific method, particularly statistical analysis, contributes to systematic analytical errors, including misuse of parametric tests, unrecognized confounding, p-hacking [1] and publication bias [2].”

“Many years ago, Sterling (1959) identified a 97% “ success ” rate across four journals in psychology, meaning that 97% of published articles reported statistically significant support of the central hypotheses.”

“The high success rate of SRs is supported by reviews of non-registered articles from different disciplines fi nding that more than 95 percent of all articles present positive fi ndings (95% for psychology journals, 85% for medical journals; Sterling et al. 1995:109; Sterling 1959:32).”

“Reducing publication bias, for example, could help correct the publication record (Nosek et al., 2018, 2019)—for early discussions, see also (Sterling, 1959) and Greenwald (1975 ).”

“Ce phénomène, détecté dès la fin des années 1950 (Sterling, 1959), constitue un biais de publication majeur, source d’au moins deux distorsions.”

“However, the scientific cycle suffers from multiple threats, including publication bias (Franco et al., 2014; Ioannidis, 2014; Sterling, 1959) and questionable research practices (see Bakker et al., 2012; John et al., 2012 for a review).”

“Sterling [4] die Konsequenzen des Wissenschaftssystems, welches lediglich Studien mit einem statistisch signifikanten Ergebnis publiziert, da die Wahrscheinlichkeit per Zufall ein signifikantes Ergebnis zu erzielen bei 5 % liegt (Kasten).”

“…resultados estatisticamente significativos ou tendentes à validação do conhecimento preexistente, constitui uma preocupação sempre presente (Card & Krueger, 1995; Doucouliagos, Laroche & Stanley, 2005; Humphreys, Reigel & Epstein, 1995; Rosenthal, 1979; Stanley, 2008; Sterling, 1959, entre outros).”

“from a beta(8, 8) distribution, given covariate x = 1; from a beta mixture 1 2 beta(5, 6) + 1 2 beta(40, 10) distribution given covariate x = 2; from an exponential Exp(8) distribution55 truncated above by 1, given x = 3; and from the beta mixture 1 2 beta(4, 4)+ 1 2 beta(150, 150) given x = 4 (distributions are from.”

“Sterling(1959),examiningarticlesinvolvinghypothesistesting 17 more pages are available in the full version of this document, which may be purchased using the "Add to Cart" button on the product's…”

“in four major psychological journals were statistically significant, thus supporting the conclusion that studies with non-significant results were underreported (Sterling, 1959).”

“Ceux-ci comprennent les études médicales et psychologiques (Sterling, 1959; Rosenthal, 1979; Begg et Berlin, 1988; Mamoulakis et al., 2009), l'organisation industrielle (Djankov et Murrell, 2002 ; Allouche et al., 2008), l'innovation et la productivité (Damanpour, 1991; Gorg et Strobl, 2001 ;…”

“Especially in statistics and psychology, scholars have always questioned the rigor and honesty of their field’s practices with every generation (e.g., Sterling 1959).”

“In addition, it has been documented that certain studies remain unpublished because of theoretical or ideological divergences, or conflicts of interest between researchers (Sterling, 1959; Mahoney, 1977).”

“In fact, it is well known that there is a large bias in science against the publication of negative results (Sterling 1959; Sterling et. al.”

“Psychology has had a replicability problem for decades (Sterling, 1959), but it seems as if the time has finally come to improve scientific practices in psychology (Schimmack, 2012).”

“…los artículos experimentales en las publicaciones más importantes usaban estadística inferencial para justificar sus conclusiones desde los datos (Sterling, 1959), y los editores de las revistas principales hicieron del test de significación un requisito para Gigerenzer, Gerd y Strum,…”

“It often leads to highly exaggerated findings being reported in the scientific literature (Sterling 1959; Rothstein, Sutton, and Borenstein 2006; Moreno et al. 2009; Carter et al. 2019).”

“result, most studies in Psychology that get published report hypotheses that “worked” (Fanelli, 2010; Sterling, 1959; Sterling, Rosenbaum, & Weinkam, 1995).”

“…published studies in the social sciences are a biased sample of the studies that are actually conducted (Bakan, 1967; McNemar, 1960; Smart, 1966; Sterling, 1959), since journals are filled with the 5% of studies that show type I errors, while approximately 95% of the unpublished studies show…”