In the SBC paper ( https://arxiv.org/abs/1804.06788 ), the ranks are defined using the binary value of “draw < true” (see page 7 in the corresponding PDF). Apparently, rstan::sbc() uses the definition “draw > true” when defining the ranks for the SBC (see https://github.com/stan-dev/rstan/blob/develop…

I think you are right that the interpretation is reversed if the location of the computed data averaged prior is not correct. For over and under dispersion I think it should not matter. Visually with the same data: [plot1] [plot2]

SBC: definition of ranks

Algorithms

Jean_Billie October 15, 2019, 10:37am 3

He already answered here.

I think the uniformity of modified rank and the faithful rank in the paper are equivalent.

2 Likes

Topic		Replies	Views
SBC with histogram/ecdf vs treating credible intervals as confidence intervals General simulation-based-calibration	6	995	March 9, 2022
Sbc histogram interpretation General simulation-based-calibration	16	1249	September 22, 2019
New preprint on model/algorithm validation with SBC - It is stronger than you thought! Publicity papers , simulation-based-calibration	0	560	November 7, 2022
Can sbc used to adjust prior? Modeling simulation-based-calibration	1	341	July 22, 2020
SBC depends on the size of data? (not only MCMC sample size) Modeling simulation-based-calibration	1	472	July 15, 2020

SBC: definition of ranks

Related Topics