4 Citations   Views   Downloads

Inferential statistics as descriptive statistics: there is no replication crisis if we don't expect replication

View preprint
“Unreplicable research is mainly a crisis of overconfidence in statistical results.” Amrhein et al (2018) on how statistical tests model both random variation & violations of model assumptions. Read at: https://t.co/qEKtFWYHoa #phdchat #openscience #replication https://t.co/Aq7It1cEZl
RT @JadePinkSameera: @learnfromerror @Lester_Domes @f2harrell @METRICStanford @StatModeling @goodmanmetrics Deborah, are u referring to thi…
@learnfromerror @Lester_Domes @f2harrell @METRICStanford @StatModeling @goodmanmetrics Deborah, are u referring to this preprint? https://t.co/i84pBtuCFD What was its thesis? @vamrhein
RT @amyalkon: Related preprint: "Inferential statistics as descriptive statistics: there is no replication crisis if we don't expect replic…
100 days ago
Related preprint: "Inferential statistics as descriptive statistics: there is no replication crisis if we don't expect replication," by @vamrhein (Valentin Amrhein) David Traffimow @Lester_Domes (Sander Greenland) https://t.co/0cK8Ct3ZcF
109 days ago
RT @jonasobleser: This is a more radical intervention (i.e., “let go of declaring significance at all”). The actual paper (Am Stat, 2019:…
RT @jonasobleser: This is a more radical intervention (i.e., “let go of declaring significance at all”). The actual paper (Am Stat, 2019:…
This is a more radical intervention (i.e., “let go of declaring significance at all”). The actual paper (Am Stat, 2019: https://t.co/FDbWEKCf4J) is IMHO one that every stats user should at least peruse. You might also want to co-sign their comment in Nature, by @vamrhein et al. https://t.co/JoOuR0YZpU
A surfeit of quotable quotes in this preprint. Particularly enjoying the "statistics...as an automated scientific decision machine" image. https://t.co/HFnjVFOto1
RT @GuyProchilo: Coffee with a side of the latest stats literature. https://t.co/jetSZInqoa Inferential statistics as descriptive statisti…
RT @GuyProchilo: Coffee with a side of the latest stats literature. https://t.co/jetSZInqoa Inferential statistics as descriptive statisti…
Coffee with a side of the latest stats literature. https://t.co/jetSZInqoa Inferential statistics as descriptive statistics - definitely worth a read. #phdchat #phd #estimation https://t.co/pF7eSMCEI5
RT @kaz_yos: "Inferential statistics as descriptive statistics: there is no replication crisis if we don't expect replication" https://t.co…
"Inferential statistics as descriptive statistics: there is no replication crisis if we don't expect replication" https://t.co/Dx89TrYFbE https://t.co/qVCUdkrusb
It is another step to make inferences about reality. This step should be informed by (not automated via) inferential statistics from multiple studies & lines of research - due to random variation and varying model assumption violations. Nice pre-print! https://t.co/QrwVz8I90g 2/2
@jd_mathbio @statsepi @bolkerb @MPKain Don't you think that P-values should be reported in a continuous fashion, whether or not one uses them for dichotomization? Perhaps you are interested in the attached paper that will be part of the forthcoming special issue of The American Statistician. https://t.co/VNayXZikC9
RT @Lester_Domes: @Abraham_RMI @TestTreatments Sorry, I see that doi is not yet live (even though proofs are done). Until then check out th…
@PhDemetri @beeonaposy @statwonk Yes! https://t.co/VNayXZikC9
RT @Lester_Domes: @Abraham_RMI @lakens @learnfromerror @stephensenn @f2harrell @MaartenvSmeden @_MiguelHernan @david_colquhoun @statsepi He…
RT @Lester_Domes: @Abraham_RMI @lakens @learnfromerror @stephensenn @f2harrell @MaartenvSmeden @_MiguelHernan @david_colquhoun @statsepi He…
@Abraham_RMI @lakens @learnfromerror @stephensenn @f2harrell @MaartenvSmeden @_MiguelHernan @david_colquhoun @statsepi Here's another view, along the lines of drop "significance" and "confidence" terminology, and look at p continuously, preferably transformed to s=-log(p) (even if design is based on a justifiable cutoff): https://t.co/goXdkOAwDM https://t.co/KvmMU7tUmv https://t.co/bAhJGZKmq9
RT @Lester_Domes: @Abraham_RMI @TestTreatments Sorry, I see that doi is not yet live (even though proofs are done). Until then check out th…
RT @Lester_Domes: @Abraham_RMI @TestTreatments Sorry, I see that doi is not yet live (even though proofs are done). Until then check out th…
145 days ago
RT @dngman: @NeuroStats @robustgar Similar thought: https://t.co/puHXBliwy1 by @vamrhein — why should inferential stats be necessarily inf…
@NeuroStats @robustgar Similar thought: https://t.co/puHXBliwy1 by @vamrhein — why should inferential stats be necessarily inferential?
NOT PEER-REVIEWED
"PeerJ Preprints" is a venue for early communication or feedback before peer review. Data may be preliminary.

Additional Information

Competing Interests

There are no competing interests.

Author Contributions

Valentin Amrhein conceived and designed the experiments, authored or reviewed drafts of the paper, approved the final draft.

David Trafimow conceived and designed the experiments, authored or reviewed drafts of the paper, approved the final draft.

Sander Greenland conceived and designed the experiments, authored or reviewed drafts of the paper, approved the final draft.

Data Deposition

The following information was supplied regarding data availability:

This is an opinion piece that did not generate any data or code.

Funding

The authors received no funding for this work.


Add your feedback

Before adding feedback, consider if it can be asked as a question instead, and if so then use the Question tab. Pointing out typos is fine, but authors are encouraged to accept only substantially helpful feedback.

Some Markdown syntax is allowed: _italic_ **bold** ^superscript^ ~subscript~ %%blockquote%% [link text](link URL)
 
By posting this you agree to PeerJ's commenting policies