The replication crisis has raised widespread concerns across scientific disciplines, prompting numerous evaluations of research practices. Researchers are refining statistical methods to enhance the accuracy and reliability of findings. A key challenge in behavioral sciences is bias in effect estimation. Underpowered studies can lead to an inflation of effect sizes in the scientific literature. Cost-effective flexible designs, such as adaptive and group sequential designs, offer promising ways to achieve greater statistical power in a feasible manner. We will further explore and introduce these designs for studies in behavioral sciences. Such designs however also influence effect sizes. For example, studies stopping early for efficacy during interim analyses often yield overestimated effects that require correction. We therefore also aim to study how different adaptive designs impact effect size estimation and the predictive value of study results.