Welcome to the OSF home for our [paper][1]: Kurz, A. S., Flynn, M. K., & Bordieri, M. J. (2016). How Bayesian estimation might improve CBS measure development: A case study with body-image flexibility in Hispanic students. *Journal of Contextual Behavioral Science, 5*, 146–153. doi: 10.1016/j.jcbs.2016.07.005 Here you'll find a preprint version of the manuscript and materials to supplement it, such as a detailed breakdown of our M*plus* code and a look of the data themselves. When we wrote our paper, we tried to balance two distinct lines of inquiry. Our main substantive goal was to assess how reasonably the [BI-AAQ](http://www.sciencedirect.com/science/article/pii/S2212144713000069) worked with our group of all-Hispanic participants, a goal that may have interested multicultural researchers or those coming from the [contextual behavioral science](https://www.researchgate.net/publication/257744287_Contextual_Behavioral_Science_Creating_a_science_more_adequate_to_the_challenge_of_the_human_condition) community. Our methodological goal was to see whether the so called Bayesian Structural Equation Modeling (BSEM) approach had advantages over traditional factor analytic approaches. Because we recognize the BSEM approach is new and perhaps somewhat confusing, we’d like to take the opportunity to link to two seminal papers on the topic [here](https://www.statmodel.com/download/BSEMv4.pdf) and [here](https://www.statmodel.com/download/BSEMRejoinder.pdf). Here's the BSEM model: ![In this model, each residual correlation is a near, but not exact, zero.][2] [1]: http://www.sciencedirect.com/science/article/pii/S2212144716300436 [2]: https://files.osf.io/v1/resources/6dnna/providers/osfstorage/5ab69f4076e58c000df89622?mode=render