While conducting social experiments, it is essential to gather accurate data. However, there are several factors that may interfere with this process and threaten the credibility of the information. First of all, the term of internal validity should be defined; researchers note that it is a component, according to which “the results are attributable to the independent variable” (Rubin & Babbie, 2016, p. 273). There are a number of elements that should be taken into account while conducting research since they are considered threats to internal validity. One of them is selection bias, which is an error that takes place in cases where the participants in the research do not represent the target population. For example, it can occur when a researcher calculates the outcomes of the trial and excludes those who decided to drop out of the experiment. However, their results should also be in the study since they represent the target population that had no benefits because of an intervention or a program discussed in the study.
In addition, while evaluating the outcomes of the social experiments, the investigators can encounter some practical pitfalls, including “fidelity of the intervention, contamination of the control condition, resistance to the case assignment protocol, client recruitment and retention” (Rubin & Babbie, 2016, p. 212). Nevertheless, there are methods that can help to avoid such mistakes. For example, it would be helpful to monitor the work of agency staff. It should be done in order to control the proper implementation of all protocols regarding the study. Moreover, conducting a pilot study is also beneficial since it can help to discover initial mistakes of the research and to mitigate them before the main one.
Reference
Rubin, A., & Babbie, E. (2016). Essential research methods for social work (4th ed.). Brooks/Cole Cengage Learning