Beyond Experiments.

Area of Science:

• Behavioral Sciences
• Causal Inference
• Research Methodology

Background:

• Experiments are often privileged for establishing causal inferences in behavioral science.
• This overvaluation leads to overuse and underappreciation of experimental limitations.
• Other research methods are frequently neglected as a result.

Purpose of the Study:

• To challenge the privileged status of experiments in behavioral science.
• To highlight the limitations of experimental research.
• To emphasize the essential role of quasi-experimental and nonexperimental methods.

Main Methods:

• Critical analysis of experimental validity (external, construct, statistical-conclusion, internal).
• Discussion of replicability and conceptual issues in simple causality.
• Examination of the contributions of quasi-experimental and nonexperimental designs.

Main Results:

• Experiments have significant limitations, including issues with validity and generalizability.
• Simple 'X causes Y' thinking is often insufficient for understanding complex phenomena.
• Quasi-experimental and nonexperimental methods offer unique insights and causal estimates.

Conclusions:

• Experiments should not be privileged over other research methods.
• A balanced research program integrates experiments with quasi-experimental and nonexperimental approaches.
• Combining diverse methodologies enhances the robustness and scope of causal understanding.