In Bayesian statistics instead of p-values some statisticians look at the Bayes
ID: 3223725 • Letter: I
Question
In Bayesian statistics instead of p-values some statisticians look at the Bayes Factor (BF). For example BF10 is the Bayes Factor in favor of the alternative hypothesis (BF01 is in favor the null). Basically BF10 indicates how many times more likely the alternative hypothesis is at explaining the data compared to the null. Even though BF is a continuous measure, everyone loves interpretative values. Some of the most common are given by Jeffreys in 1961. Using his guidelines if you conducted a Bayesian test and found that BF10 = 16, what level of evidence would this be for the alternative against the null?
A) No evidence
B) Anecdotal evidence for H1
C) Moderate evidence for H1
D) Strong evidence for H1
E) Very strong evidence for H1
F) Extreme evidence for H1
A) No evidence
B) Anecdotal evidence for H1
C) Moderate evidence for H1
D) Strong evidence for H1
E) Very strong evidence for H1
F) Extreme evidence for H1
Explanation / Answer
Using his guidelines if you conducted a Bayesian test and found that BF10 = 16, what level of evidence would this be for the alternative against the null?
E) Very strong evidence for H1
E) Very strong evidence for H1