Well-planned empirical studies are central for testing hypotheses in the quantitative social sciences. However, the limited availability of research resources and the ambiguity of statistical decisions pose important challenges to the informativeness of research results. This dissertation outlines new avenues for increasing the informativeness and efficiency of research in psychological science. It puts an emphasis on Bayesian statistical methods and delineates how research design can benefit from the distinct features of Bayesian statistics. Specifically, it broadens the traditional view on research informativeness towards the quantification of evidence, informative model comparisons, and the mitigation of analytic flexibility.