One of the most exciting, yet frustrating, aspects of being a clinical researcher is that as we therapists clinically morph and grow, so do the statistical analyses that we use. They have to change in order to account for the complexity of the work we are doing. Simple bi-variate correlations, once sufficient, paved the way for multiple regression where confounding phenomena could be controlled for. In clinical research, this allowed researchers to account for factors such as diagnosis or duration of pre-treatment impairment when examining clinical outcomes. Then, structural equation modeling (SEM) gained popularity as a tool beyond multiple regression, an improved approach for a number of reasons. SEM allowed researchers to account for measurement error, to incorporate multiple dependent variables into their research questions, to test how well a model fits the data, and to work with complex and difficult data (Wang & Wang, 2012). This allowed clinicians to account for complex change processes happening over the course of treatment. Nothing that we do clinically is simple, so why should the analyses we use to validate our approach be? Standards of practice for running and reporting statistical analyses continued to evolve. Testing and reporting of statistical power, effect sizes, and confidence intervals have increasingly become critical components of the research process to manage typical errors that arise from an overreliance on the p < .05 criterion (Cohen, 1994; Cumming, 2012). Other complex approaches emerged as means to manage the complexity of the researcher and the data. Meta-analysis became the new standard for reviewing clinical studies because of its ability to reduce researcher bias when synthesizing the results of studies. Dyadic analysis approaches such as the Actor-Partner Interdependence Model (APIM, Cook & Snyder, 2005) opened the door to more appropriate and valid analyses of data that included paired participants, such as romantic partners and parent-child dyads. These changes are exciting because they enable us to improve the precision and accuracy of our statistical analyses, thereby improving the validity of our findings (Johnson & Miller, 2014). Moreover, it is exciting to have new statistical approaches that can better model the multiple relationships and complex interactions that are inherent in couple and family therapy. However, they are also frustrating because the standard for doing state-of-the-art statistics seems to always be increasing. Learning a new statistical approach does not bring much satisfaction because, as soon as we learn it, we realize that there is something new that we need to learn. The frustration comes from feeling as if our statistical knowledge is perpetually outdated. And, make no mistake about it; newer statistical techniques seem to gravitate, like some law of physics, toward greater complexity and sophistication. Well, the evolution has continued, and there is a new generation of statistical approaches that are available to CFT researchers to enhance their ...