ABSTRACT
An experiment and a survey are alternative methods of getting knowledge about the relationships between variables, most especially about the causal relationships between variables. The principle of experimenting when possible seems obvious in the late twentieth century. Deciding how many variables with which to experiment is also an important decision. In an experiment the researcher purposely alters the environment in order to observe the effect of the experimentally altered conditions. Medicines are also always tested in real-life experiments before they are released for general use, but only after non-real-life experiments indicate that the real-life experiments probably will not be dangerous. The number of experimental groups depends upon how many different stimuli researchers want to test or on how many different levels of the independent variable they want to test. The most important advantage of the experiment is that the relationship that researchers actually observe is clear in its causal direction.
