Behavioral economic research increasingly faces criticism: effects may not replicate, scale outside the laboratory, or last over an extended period. The field’s focus is now on understanding not only what works but also when and why. Carefully designed experiments can help identify boundary conditions of behavioral interventions. In this thesis, I present three experimental studies that investigate human behavior. Each of the studies contributes to the scholarly discussion with its research design either emphasizing mechanisms (the why) or identifying intervention effects outside the laboratory or in the long run (what works/when). In the first chapter, we examine gender differences in claiming credit for contributions to teamwork. Data from a laboratory experiment show that both men and women slightly overestimate their contributions, but men are more willing to exaggerate when it pays off, even if it harms another team member. In the second chapter, we evaluate the effectiveness of several different theory-based interventions to prevent littering in social housing over a period of seven weeks. The field experiment shows that posters of watching eyes and untouched nature work better than information on monetary costs and norms. In the third chapter, I report results from a large-scale lab-in-the-field experiment to test whether thinking about a positive future event (“episodic future thinking”) induces precautionary saving among adolescents. I find that this is not the case and discuss the robustness of this result.