Title: In Defense of the Flat Prior, at Least for SEM Estimates and Inferences Author: Timo von Oertzen Affiliation: Universität der Bundeswehr, München Abstract: Bayesian statistics comes with two main advantages for social scientists, the possibility to specify prior knowledge and the interpretability of the posterior knowledge. Sadly, the second comes with a price for the first: Since most social scientist are not able (both for substantive and technical reasons) to provide a subjective prior, they must resort to an uninformative one, which again usually comes at the expense of not really understanding the posterior probability. In this presentation, we will discuss why these expenses are minimal for a flat prior. However, many good reasons have been given why a flat prior is a questionable choice. We will also discuss why these disadvantages are not as severe as they may seem at first glance in the context of modern social data analysis.