tag:blogger.com,1999:blog-3240271627873788873.post2715963128415224695..comments2018-05-21T08:11:58.507-04:00Comments on Doing Bayesian Data Analysis: Just published: "Rejecting or Accepting Parameter Values in Bayesian Estimation"John K. Kruschkehttp://www.blogger.com/profile/17323153789716653784noreply@blogger.comBlogger6125tag:blogger.com,1999:blog-3240271627873788873.post-41152322632444885232018-05-18T00:33:09.775-04:002018-05-18T00:33:09.775-04:00That is the discussion I was looking for, thanks.
...That is the discussion I was looking for, thanks.<br /><br />The HDI+ROPE approach is probably better protection against the kinds of erroneous interpretations you describe, if you don't bother to look at the posterior. But I don't think I'd ever consider interpreting either result (HDI+ROPE or ROPE-only) without actually looking at the posterior distribution.Serje Robidouxhttps://www.blogger.com/profile/11056504396794376624noreply@blogger.comtag:blogger.com,1999:blog-3240271627873788873.post-78910772218857004042018-05-17T20:38:32.151-04:002018-05-17T20:38:32.151-04:00In the Supplement, at https://osf.io/fchdr/ , ther...In the Supplement, at https://osf.io/fchdr/ , there's a section headed, "Decision rule based on ROPE alone." That might address your question. If not, let me know.John K. Kruschkehttps://www.blogger.com/profile/17323153789716653784noreply@blogger.comtag:blogger.com,1999:blog-3240271627873788873.post-9790793107178343812018-05-17T19:07:47.995-04:002018-05-17T19:07:47.995-04:00Really nice, simple to understand paper.
There...Really nice, simple to understand paper.<br /><br />There's one thing that I'm not sure I can get my head around:<br />Is there a particular reason for using the location of the HDI rather than simply asking how much of the distribution falls above vs below the upper bound of the ROPE? Why ignore the density in the outer tails when making the decision?Serje Robidouxhttps://www.blogger.com/profile/11056504396794376624noreply@blogger.comtag:blogger.com,1999:blog-3240271627873788873.post-4817903479240435502018-05-13T08:31:37.029-04:002018-05-13T08:31:37.029-04:00P.S. If you haven't seen it, be sure to read t...P.S. If you haven't seen it, be sure to read this blog post about the values of nu: https://doingbayesiandataanalysis.blogspot.com/2015/12/prior-on-df-normality-parameter-in-t.html<br />John K. Kruschkehttps://www.blogger.com/profile/17323153789716653784noreply@blogger.comtag:blogger.com,1999:blog-3240271627873788873.post-92046461406921970862018-05-13T08:29:54.668-04:002018-05-13T08:29:54.668-04:00It would be straight forward to give each group it...It would be straight forward to give each group its own nu, analogous to giving each group its own sigma. However, the value of nu is influenced mainly by the outlying data, and typically each group has relatively small numbers of outlying data. Therefore the estimate of nu is typically very uncertain for each group. By using one nu for all groups, we are using the outliers from all groups to inform the tail-heaviness of the noise distribution. Using one nu for all groups might be interpreted as assuming that the same outlier process affects all groups (whatever that means). If you have a situation in which you have a ton of data in the tails and reason to believe that the normalities of the groups differ, then do give each group its own nu parameter.John K. Kruschkehttps://www.blogger.com/profile/17323153789716653784noreply@blogger.comtag:blogger.com,1999:blog-3240271627873788873.post-18224112525292251602018-05-12T21:51:21.668-04:002018-05-12T21:51:21.668-04:00Hello, Mr. Kruschke, good night.
In your robust m...Hello, Mr. Kruschke, good night.<br /><br />In your robust models, I see you use an individual mu and sigma for each distribution, but always a shared nu for both. Why not an individual nu for each distribution? Are there cases when one should use an individual nu for each distribution?<br /><br />Thanks in advance.Heitor Thuryhttps://www.blogger.com/profile/06808665933888019334noreply@blogger.com