I spent some time today trying to see what phoneticians are doing in reporting statistical effects using Bayesian analysis. Two perennial questions: (1) do I report just my stats; or the data & the stats? (2) do figures include the posterior distribution, the data, or both? 1/6

So, in @echodroff & Wilson 2019,
-The figures are the data, not the model.
-They list beta coefficients with 95% confidence intervals, which I think is probably a direct report from the hypothesis function. 2/6 https://link.springer.com/article/10.3758/s13414-019-01894-2">https://link.springer.com/article/1...

In @jvcasill (Casillas) 2019 (ICPhS),
-The figures are parameter estimates alongside a plot of the confidence intervals (so plotting the posterior).
-But figure 3 is the data distribution alongside posterior probability distribution. 3/6

https://assta.org/proceedings/ICPhS2019Microsite/pdf/full-paper_899.pdf">https://assta.org/proceedin...

In Seyfarth et al. (2019),
-They report the median estimate and Bayes factor range, which appears to be slightly different from the studies above.
-There are no figures, but only model output seems to be reported (no raw/normalized data). 4/6

https://link-springer-com.gate.lib.buffalo.edu/content/pdf/10.3758/s13423-019-01637-2.pdf">https://link-springer-com.gate.lib.buffalo.edu/content/p...

In Kim & @katiedrager (2018),
-They report a table with mean values, sd, CI intervals, and a p value (perhaps by querying the probability of the posterior samples, as per Franke & @TimoRoettger?)
-The figures are the data, not the model. 5/6

https://onlinelibrary.wiley.com/doi/pdf/10.1111/tops.12351">https://onlinelibrary.wiley.com/doi/pdf/1...

That& #39;s my reading of current practices at least. My take-away is that I don& #39;t need to worry about plotting a posterior distribution per se (I think) even though it seems like all the tutorials on brms do this. Apologies for any errors above - I& #39;m a novice with this approach.

6/6