By manipulating the relevant independent variable with accuracy and reducing any bias by assigning groups/treatments randomly, this allows for any related change in the dependent variable to be observed. If this is done (experimentally) in a laboratory where control over this process is high and the results measured are relatively significant, the researcher can with a high level of confidence conclude a relationship of the nature "if x, then y". Therefore, the degree of certainty that the laboratory experimenter enjoys in his/her conclusion is due to the high level of control that s/he can exercise through experimental manipulation of the variables in the environment. However, in a laboratory, the extent to which these variables can be manipulated is sometimes limited, and this requires other methods. .
Field experiments have the advantage of magnifying the influence of the independent variable in a present relationship of for example a societal level phenomenon. A variable such as "group pressure" grows as the group grows and naturally therefore, when moving from a laboratory to a field, this will have a magnifying effect on such variable. This could allow the investigated relationship to become more visible and if carried out carefully could also contribute greatly to the external validity of the found results, as the experiment is carried out in a natural setting. Unfortunately, in practise, this method has its trade offs as it looses the degree of control that a laboratory experimenter can achieve. Therefore, although a relationship is now more visible due to the magnification that a field experiment provides, the proof of the exact nature of the found relationship becomes less convincing. Because of this lack of control, one of the preoccupations of the field experimenter should be to make the conditions approximate a laboratory experiment as much as possible (Kerlinger and Lee, 2000 p.
