Experimental method is a method in which a variable (independent variable that is hypothesized as a cause; IV) is manipulated by an experimenter and the corresponding change in another variable (dependent variable that is hypothesized as an effect; DV) is observed. To determine whether the change in the DV is caused by the IV, at least two groups are involved: a control group and an experimental group. The two groups are assumed to be identical in all respects except that the control group does not receive the treatment of the IV whereas the experimental group receives the treatment. In reality, no two groups are exactly identical. Therefore, to ensure both groups are identical except for the treatment or experimental manipulation of the IV, all participants are randomly assigned to either the control or experimental group(s). As a result, any innate differences between the members of the two groups are equally distributed.
As an example, consider a hypothetical study investigating the effect of violent TV programs on the behavior of children. The researcher first randomly assigns a group of children to either an experimental or control group, and then shows a violent TV program to children in the experimental group and a neutral program to children in the control group. After viewing their programs, the children are allowed to play in a room with other children and observed. As a result, children in the experimental condition may exhibit more aggressive behavior than children in the control group. Given the use of a control group and an experimental group, and the random assignment of children to each condition, the researcher may be able to infer that the violent TV program caused aggression among the children since they only differed in the type of program they watched.
Experimental methods facilitate making causal inferences easier. In a well-designed experiment, the relationship between the IV and DV is clear. The experimental method can isolate the relationship among variables, which occurs in a complex environment and is often unobservable. As in the example, the effect of violent TV programs is not easily observed in real settings. By controlling all of the other variables involved, a well-designed experiment makes it possible to discern the relationship between the IV(s) and DV(s). By doing so, experiments can provide grounds for further scientific investigation. In other situations, the relationship between any two variables is weak. Because experimenters can have a lot of control in an experiment, they may maximize the magnitude of the manipulation and thereby have higher power to detect the relationship. In these scenarios, the major concern of the experimenter is the existence of the relationship. Especially at an exploratory step, this kind of evidence may provide a clue as to whether further investigation is warranted.
The experimental method is not without its shortcomings. First of all, its major advantage can often be a disadvantage. An experimenter’s control over many aspects of the experiment often makes it hard to generalize the results to other situations. Therefore, the size of the effect in an experiment may not be observed in reality or in other studies. This is intensified because many variables are intertwined with other variables. In this sense, experiments are much simpler than reality. Experimenters may try to include more variables in the design to increase the generalizability of the results; however, often the inclusion of additional variables poses a challenge to experimenters. Experiments typically require more time and effort for each participant. This might be a part of the reason why experiments are not used to collect longitudinal data. Maintaining control over participants for long periods of time may cause ethical issues, and may even be impossible.
References:
- Pedhazur, E. J., & Schmelkin, L. P. (1991). Measurement, design, and analysis: An integrated approac Hillsdale, NJ: Erlbaum.
- Tabachnick, B. G., & Fidell, L. S. (2001). Computer-assisted research design and analysis. Needham Heights, MA: Allyn &
- Trochim, W. (2002). Experimental design. Retrieved from http://trochim.human.cornell.edu/kb/desexper.htm
- Whitley, B. E., Jr. (1996). Principles of research in behavioral science. Mountain View, CA: Mayf
- Wolfe, L. M. (n.d.). Developmental research methods.
- Retrieved from http://www.webster.edu/~woolflm/methods/dehtml