Data-dependent analysis—a "garden of forking paths"— explains why many statistically significant comparisons don't hold up.

130 citations recorded by the source corpus.

How researchers see this paper

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

Innovation and profit motivations for social entrepreneurship: A fuzzy-set analysis (2019, methodology): “Greckhamer et al. (2018) stress that all conditions included in the model should have a theoretical or strong empirical rationale for being included, since although fsQCA is an inductive, iterative method, it is not a “fishing expedition” (Gelman & Loken, 2014).”
Visualization in Bayesian workflow (2017, methodology): “A different objection to using the data twice (or even more) comes from ideas around hypothesis testing and unbiased estimation, but we are of the opinion that the danger of overfitting the data is much more concerning (Gelman and Loken, 2014).”
The influence of hidden researcher decisions in applied microeconomics (2020, unknown): “…in the process even if analysis does not change after viewing the results, so the issue is distinct from “p-hacking” or “The Garden of Forking Paths” (Gelman & Loken, 2014; Silberzahn et al., 2018), as well as being distinct from publication bias (De Long & Lang, 1992), and would not be solved…”
Effectiveness of spelling interventions for learners with dyslexia: A meta-analysis and systematic review (2020, background): “It is well known that publication bias and questionable research practices inflate effect size estimates of meta-analyses (Gelman & Carlin, 2014; Gelman & Loken, 2014; Schmidt, 1992).”