The earth is flat (p > 0.05): significance thresholds and the crisis of unreplicable research

View article
RT @vamrhein: The null hypothesis cannot be 'confirmed' nor 'strengthened' using a p-value, because very likely there are many better hypot…
reply repost on X like
RT @vamrhein: The null hypothesis cannot be 'confirmed' nor 'strengthened' using a p-value, because very likely there are many better hypot…
reply repost on X like
RT @vamrhein: The null hypothesis cannot be 'confirmed' nor 'strengthened' using a p-value, because very likely there are many better hypot…
reply repost on X like
The earth is flat (p>0.05) https://t.co/3Olio08Rz4 via @thePeerJ
reply repost on X like
RT @vamrhein: The null hypothesis cannot be 'confirmed' nor 'strengthened' using a p-value, because very likely there are many better hypot…
reply repost on X like
RT @vamrhein: The null hypothesis cannot be 'confirmed' nor 'strengthened' using a p-value, because very likely there are many better hypot…
reply repost on X like
p-values tell little about #reliability of #research https://t.co/LZztTW46LO #epitwitter #epidemiology
reply repost on X like
One more about significance & replicability https://t.co/LZztTW46LO #epitwitter #epidemiology
reply repost on X like
RT @vamrhein: The null hypothesis cannot be 'confirmed' nor 'strengthened' using a p-value, because very likely there are many better hypot…
reply repost on X like
2584 days ago
RT @vamrhein: The null hypothesis cannot be 'confirmed' nor 'strengthened' using a p-value, because very likely there are many better hypot…
reply repost on X like
RT @vamrhein: The null hypothesis cannot be 'confirmed' nor 'strengthened' using a p-value, because very likely there are many better hypot…
reply repost on X like
RT @vamrhein: The null hypothesis cannot be 'confirmed' nor 'strengthened' using a p-value, because very likely there are many better hypot…
reply repost on X like
RT @vamrhein: The null hypothesis cannot be 'confirmed' nor 'strengthened' using a p-value, because very likely there are many better hypot…
reply repost on X like
RT @vamrhein: The null hypothesis cannot be 'confirmed' nor 'strengthened' using a p-value, because very likely there are many better hypot…
reply repost on X like
The null hypothesis cannot be 'confirmed' nor 'strengthened' using a p-value, because very likely there are many better hypotheses: Any p-value less than 1 implies that the null hypothesis is not the hypothesis most compatible with the data. https://t.co/1Ys3mYi6Df
reply repost on X like
RT @vamrhein: Whenever we use effect sizes, confidence intervals, AIC, posteriors, Bayes factors, likelihood ratios, or false discovery rat…
reply repost on X like
2585 days ago
RT @vamrhein: Whenever we use effect sizes, confidence intervals, AIC, posteriors, Bayes factors, likelihood ratios, or false discovery rat…
reply repost on X like
2585 days ago
RT @vamrhein: Within a 95% compatibility interval, the sample average is about seven times as plausible, or seven times as good a bet for t…
reply repost on X like
2586 days ago
RT @danilobzdok: Inspiring take on the relation between #pvalues and #reproducibility: "two studies will be ‘conflicting’, meaning that on…
reply repost on X like
RT @danilobzdok: Inspiring take on the relation between #pvalues and #reproducibility: "two studies will be ‘conflicting’, meaning that on…
reply repost on X like
RT @danilobzdok: Inspiring take on the relation between #pvalues and #reproducibility: "two studies will be ‘conflicting’, meaning that on…
reply repost on X like
2587 days ago
RT @danilobzdok: Inspiring take on the relation between #pvalues and #reproducibility: "two studies will be ‘conflicting’, meaning that on…
reply repost on X like
Within a 95% compatibility interval, the sample average is about seven times as plausible, or seven times as good a bet for the true population average, as are the limits of the interval (given our model is correct). https://t.co/tGsxj3AGXQ
reply repost on X like
Inspiring take on the relation between #pvalues and #reproducibility: "two studies will be ‘conflicting’, meaning that one is significant and the other is not, in one third of the cases if there is a true effect" https://t.co/COvaRBiACE @talyarkoni @russpoldrack via @dngman https://t.co/QInXe2DyWI
reply repost on X like
RT @vamrhein: Quite often, significance testing appears like a sort of gambling. And we may have become somewhat addicted to this game—it i…
reply repost on X like