A surge of p-values between 0.041 and 0.049 in recent decades (but negative results are increasing rapidly too)

View article
RT @pmgjones: Hmm. Anything seem amiss here? P-Hacking Is definitely occurring in the biomedical literature. When are we going to have the…
reply repost on X like
RT @pmgjones: Hmm. Anything seem amiss here? P-Hacking Is definitely occurring in the biomedical literature. When are we going to have the…
reply repost on X like
RT @pmgjones: Hmm. Anything seem amiss here? P-Hacking Is definitely occurring in the biomedical literature. When are we going to have the…
reply repost on X like
2320 days ago
RT @pmgjones: Hmm. Anything seem amiss here? P-Hacking Is definitely occurring in the biomedical literature. When are we going to have the…
reply repost on X like
RT @pmgjones: Hmm. Anything seem amiss here? P-Hacking Is definitely occurring in the biomedical literature. When are we going to have the…
reply repost on X like
2320 days ago
RT @pmgjones: Hmm. Anything seem amiss here? P-Hacking Is definitely occurring in the biomedical literature. When are we going to have the…
reply repost on X like
RT @pmgjones: Hmm. Anything seem amiss here? P-Hacking Is definitely occurring in the biomedical literature. When are we going to have the…
reply repost on X like
2321 days ago
Hmm. Anything seem amiss here? P-Hacking Is definitely occurring in the biomedical literature. When are we going to have the adult conversation about how universities, with their focus on numbers of publications (rather than quality), are driving this? https://t.co/wLNmuXvOnN https://t.co/KR7S2YitZi
reply repost on X like
2355 days ago
Surge of pvalues between 0.041 & 0.049 in recent decades (but negative results are increasing too) https://t.co/srrIfyCa07 https://t.co/HDdK3F4DC7
reply repost on X like
RT @stephenfloor: @arjunrajlab p < 0.05 doesn't indicate truth just like a two-fold change doesn't indicate importance. Since we often use…
reply repost on X like
2482 days ago
RT @stephenfloor: @arjunrajlab p < 0.05 doesn't indicate truth just like a two-fold change doesn't indicate importance. Since we often use…
reply repost on X like
2482 days ago
RT @stephenfloor: @arjunrajlab p < 0.05 doesn't indicate truth just like a two-fold change doesn't indicate importance. Since we often use…
reply repost on X like
@arjunrajlab p < 0.05 doesn't indicate truth just like a two-fold change doesn't indicate importance. Since we often use 0.05 threshold we see effects as in the article below - would surely see the same if we settled on a threshold for "important" based on effect sizes https://t.co/akHcbBCnjF
reply repost on X like
“P-values between .041 & .049 have increased to a greater extent than those between .051 to .059 (ratio = 3.6)” de Winter and Dodou’s findings suggest that questionable research practices have increased over the past 25 years Link: https://t.co/6l7n3V1rSh #phdchat #openscience https://t.co/ncajYAqLmh
reply repost on X like
"The underlying causes of [questionable research practices/p-hacking] have purported to be an emphasis on productivity, high rejection rates of journals, and competitive schemes for funding and promotion." Read more in de Winter and Dodou (2015) https://t.co/6l7n3V1rSh #phdchat https://t.co/8gx8xyFLZH
reply repost on X like
RT @GuyProchilo: “It is well known that p values just below .05 are more prevalent than just above .05. One explanation is that many findin…
reply repost on X like
“It is well known that p values just below .05 are more prevalent than just above .05. One explanation is that many findings in the literature are false positives”. Read more in de Winter and Dodou’s (2015) survey on near-threshold p values. https://t.co/6l7n3V1rSh #phdchat https://t.co/8G2EQqjYN3
reply repost on X like
RT @GuyProchilo: “P-values between .041 & .049 were 65.7 times more prevalent in the biological sciences than the physical sciences” result…
reply repost on X like
“P-values between .041 & .049 were 65.7 times more prevalent in the biological sciences than the physical sciences” results of de Winter and Dodou’s (2015) survey on the prevalence of near-threshold results across the sciences Link: https://t.co/6l7n3V1rSh #phdchat #openscience https://t.co/mC5knTYFQT
reply repost on X like
RT @matkomorowski: A surge in p-values between 0.041 and 0.049 in the last decades!!! p-hacking or publication bias ? #Statistics #research…
reply repost on X like
2768 days ago
RT @matkomorowski: A surge in p-values between 0.041 and 0.049 in the last decades!!! p-hacking or publication bias ? #Statistics #research…
reply repost on X like
RT @matkomorowski: A surge in p-values between 0.041 and 0.049 in the last decades!!! p-hacking or publication bias ? #Statistics #research…
reply repost on X like
RT @matkomorowski: A surge in p-values between 0.041 and 0.049 in the last decades!!! p-hacking or publication bias ? #Statistics #research…
reply repost on X like
RT @matkomorowski: A surge in p-values between 0.041 and 0.049 in the last decades!!! p-hacking or publication bias ? #Statistics #research…
reply repost on X like
RT @matkomorowski: A surge in p-values between 0.041 and 0.049 in the last decades!!! p-hacking or publication bias ? #Statistics #research…
reply repost on X like