Laboratory Experiments

Purpose and Function of Laboratory Experiments

Although Wilhelm Wundt (1832–1920) is credited with developing the first laboratory for psychological experiments at the University of Leipzig in 1879, use of laboratories for social scientific research started to blossom following World War II. Some of the early experiments in communication studies were developed to explore the persuasive nature of propaganda. Laboratory experiments in communication have grown to include empirical research on media effects, health behavior, and biophysical associations with communicative phenomena, among others. Regardless of the research topic, the purpose of laboratory experiments is to test theoretical assumptions by focusing on one or several predictive elements of the theory and isolating the influence of the hypothesized elements on outcomes. Laboratory experiments play an important role in the scientific method of building, testing, and revising theories. Researchers do not necessarily use laboratory experiments to generalize to real-world phenomena occurring in a larger population, but instead they use theory-driven experiments to test hypotheses that provide [Page 836]evidence for a predicted relationship between variables, rule out alternative explanations, and lead to inferences about cause and effect relationships. In other words, laboratory experiments are not environments to build theories but to test them under specific conditions.

The main function of laboratory experiments is controllability. Research studies conducted in naturally occurring environments (e.g., surveys, content analysis, observational studies, and field experiments to a lesser extent) are more susceptible to confounding variables influencing the relationships between independent and dependent variables and other threats to internal validity. Laboratory experiments reduce such risks through careful design, thereby offering more explanatory power to show specific influences of predictor variables on outcomes. Although the behavior of a single individual is difficult to predict, experimenters assume that in highly controlled environments behavioral patterns of aggregates (i.e., groups of people) can be discovered and predicted. In this sense, laboratory experiments offer the best way to test specific behaviors directly leading to specific outcomes and what outcomes will occur under what conditions. Ideally, researchers would use laboratory experiments to account for all potential factors influencing a particular outcome, but this is practically impossible, especially in studies investigating human behavior and communication. However, researchers use several strategies in laboratory experiments to control the influence of unmeasured variables and create opportunities for the full expression of tested variables.

...