F1000Prime recommendation of False-positive psychology: undisclosed flexibility in data collection and analysis allows presenting anything as significant.

By H. V. Beek. 2012

1,005 citations recorded by the source corpus.

View Paper: https://www.semanticscholar.org/paper/6d4dc4d75d7c3cb9ed193e171389963e5737d8db

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Statistics and Probability

Citations in the corpus

Similarity-based interference in sentence comprehension: Literature review and Bayesian meta-analysis (2017, methodology): “These are shown in Example 1, and are taken from Sturt (2003). We will use this example to introduce key terminology that is used in the present paper; a summary of the terms appears in Table 1.”
Women's Representation in Science Predicts National Gender-Science Stereotypes: Evidence from 66 Nations. (2015, methodology): “Rather than using a single criterion, we report results across many choices of selection criteria, as advocated by Simmons, Nelson, and Simonsohn (2011).”
The Rules of the Game Called Psychological Science (2012, background): “Simmons et al. (2011) illustrated how easy it is to inflate Type I error rates when researchers employ hidden degrees of freedom in their analyses and design of studies (e.g., selecting the most desirable outcomes, letting the sample size depend on results of significance tests).”
Controlling Bias in Adaptive Data Analysis Using Information Theory (2015, background): “It is commonly recognized that such adaptivity (also called researcher degrees of freedom), even if well-intentioned, can lead to false discoveries, contributing to the crisis of reproducibility in science.”
Using independent covariates in experimental designs: Quantifying the trade-off between power boost and Type I error inflation (2017, methodology): “In contrast, in our simulations, flexibly including a single, carefully chosen covariate when testing the effect of X on Y inflated Type I error rates to only 6–8%, while offering a power boost comparable to that of Simmons et al.’s four-part approach. 3. Resist the temptation to toss a covariate in”
How Bayes factors change scientific practice (2016, background): “That is, bias introduced by the judicious dropping of conditions, discussed by Simmons et al. (2011), is not in itself solved by using Bayesianmethods on the data that remains.”
Small Sample Size Solutions (2020, background): “Thus, researcher degrees of freedom and publication bias can lead to overestimation of effects and an inflated false-positive rate in the literature (Simmons et al., 2011).”
Max-Information, Differential Privacy, and Post-selection Hypothesis Testing (2016, background): “If φi = A(X), then rejecting the null hypothesis when pi(φi(X)) ≤ α may result in a false discovery with probability much larger than α (indeed, this kind of naive approach to post-selection inference is suspected to be a primary culprit behind the prevalence of false discovery in empirical science ”
Meta-Analysis of Menstrual Cycle Effects on Women’s Mate Preferences (2014, background): “Effects decline over time when initial estimates are false positives that subsequently fail to replicate (Simmons et al., 2011).”