The Traditional Experimental Designs
The traditional experimental designs are used to determine whether there is a causal relationship or whether an independent variable influences the dependent variable. The elements include the use of independent and dependent variables, where the former produces a change in the latter, such as an increase or decrease (Gabrenya, 2003).
Pretesting and Post-testing
Pretesting involves measuring the subject of the experiment before it begins, whereas post-testing involves assessing the subjects after the experiment is completed. An experimental group is one that is exposed to the experimental stimulus or conditions, but the control group is not. Double-blind experiment refers to the situation in which the subjects and experimenters are not aware whether they are in the experimental or control group. Randomization entails assignment of random subjects to control or experimental groups.
The Classical Experimental Design
The classical experimental design is suitable for establishing causal relationship since it allows for comparison of an experimental group and the control group. One variable (independent) is manipulated by exposing it to the experimental conditions and finding the effect it has on another (dependent).
Classical experimental design is suitable for studying real-life situations such as the impact of viewing a certain television program on certain behaviors. For instance, it can be used to determine whether watching a film focusing on women in politics has an impact on prejudice against women. In this case, prejudice represents the dependent variable while watching the film or not represents the independent variable. The subjects under study are divided into experimental that is exposed to the stimulus and control groups that do not receive the stimulus (Bhattacherjee, 2012). Levels of prejudice on women are measured before and after the experiment after which conclusions are made.
Quasi-experiments
Quasi-experiments are used when researchers are interested in independent variables, where randomization is not possible, especially if the independent variable represents an innate characteristic such as the level of happiness (Bhattacherjee, 2012). For instance, when doing an education experiment, a researcher can divide the class alphabetically or by seating arrangement.
Correlation Experiment
Correlation experiment establishes a relation between two variables but does not provide proof that one variable can lead to a change in the other (Gabrenya, 2003). For instance, in research, one may ask parents to give the amount of violent television programs watched by their children per week. The research can monitor the behavior of the children and establish the level of aggression depending on the number of hours a child watches such programs in a week.
References
Bhattacherjee, A. (2012). Social science research: Principles, methods, and practices. http://scholarcommons.usf.edu/cgi/viewcontent.cgi?article=1002&context=oa_textbooks
Gabrenya, W.K (2003). Overview of Psychology Research Methods. Research Skills for Psychology Majors: Everything you need to know to get started. http://my.fit.edu/~gabrenya/IntroMethods/eBook/methods.pdf
