#statstab #492 A tiny review on e-values and e-processes
Thoughts: How to be a bayesian while wearing a frequentist hat.
#evalues #eprocess #samplesize #evidence #power #sequential #error #type1
www.math.uwaterloo.ca/~wang/files/...
Latest posts tagged with #evalues on Bluesky
#statstab #492 A tiny review on e-values and e-processes
Thoughts: How to be a bayesian while wearing a frequentist hat.
#evalues #eprocess #samplesize #evidence #power #sequential #error #type1
www.math.uwaterloo.ca/~wang/files/...
#statstab #376 Safe Flexible Hypothesis Tests for Practical Scenarios
Thoughts: e-values & safe tests "conserve the type I error guarantee (false positive rate) regardless of the sample size"
#SafeAnytimeValidInference #SAVI #evalues #evidence
#safetests
cran.r-project.org/web/packages...
Just gave my talk at the @KingstonUni Psychology Research Conference! Sharing insights on e-values and Safe Anytime Valid Inferences (SAVI)βpowerful tools for making statistical inferences that remain valid no matter when you stop collecting data.
#conference #evalues #savi
#statstab #308 A Tutorial on Safe Anytime-Valid Inference [t-test]
Thoughts: Safe Anytime Valid Inferences (SAVI) are the future for frequentist stats. Since #253 I'm more convinced.
#SafeAnytimeValidInference #SAVI #evalues #likelihood #evidence #ttest
osf.io/mdbqe/
Top academics have shifted their attention to e-values, Vovk was one of the first researches in the domain, other high profile researchers are picking the topic as well now.
#evalues #conformalprediction
#statstab #253 Anytime-valid inference in N-of-1 trials
Thoughts: Are frequentists ready to talk about evidence? Safe Anytime Valid Inferences (SAVI) seem like the future.
#SafeAnytimeValidInference #SAVI #evalues #likelihood #evidence
arxiv.org/abs/2309.07353
{first of 2025 π₯³}
#statstab #251 E-values
Thoughts: These seem like the impossible; they are generalizations of likelihood ratios (i.e., used for composite hypotheses), and can provide evidence for/against a hypothesis.
#evalues #evidence #likelihood
sas.uwaterloo.ca/~wang/files/...