Читать книгу Social Work Research Methods - Reginald O. York - Страница 113

It is easy to make the mistake of assuming that you are not permitted, under any circumstances, to create any risk to the human subject of research. Can you, for example, have someone fake a heart attack on a busy city street so that you can observe the behavior of people on the street who observe this behavior? Maybe you want to know how many will help, or how they react emotionally. Can you do this? The answer depends on the level of the risk and the level of the benefits of the study. If the study is viewed by your IRB as having no benefits, the board is likely to deny your proposal. If they view it as having a lot of benefits, they may approve on the basis that the benefits outweigh the risks. How likely are people on the street to have a major problem when they observe such behavior? The question here is not whether it is possible that someone may be damaged by this action but whether it is probable that people will be damaged by it. In the final analysis, it is the judgment of the IRB that will determine the outcome of the request to undertake this study. There is no formula that will guide them, only their judgment.

Risks and benefits must be addressed as an issue when you have the opportunity to deny service to a group of potential clients for a given service in order to use this group as your control group in your research study. Suppose that you have 40 persons who have asked for your special tutoring program for at-risk middle school students, and you are considering the option of randomly selecting 20 of these students to be your experimental group and randomly selecting the other 20 to be your control group. You will give your tutoring program to those in the experimental group but not to those in the control group. You will measure the students in each group to determine if the experimental group had better gains in functioning than the control group.

Is this ethical? The answer lies in the examination of risks and benefits. If you only have the resources to provide your tutoring to 20 students for the current service period, the ethical challenge is reduced. If you can only serve 20, you could select your 20 at random and use the others as the control group. If you have the resources to provide tutoring to the 20 in your control group at a later time period, the burden is further diminished. If you have the capacity, however, to serve all but will only serve 20 for the sake of research, the burden on you to justify this decision on the basis of benefit is great. You will need to make the case that this is a situation where the benefit is so great as to offset the harm to those who have been denied this service for the sake of research. Social workers seldom make such claims because of the duty to serve those in need.

The typical response of the social worker in this situation is to provide the service to all prospective clients and measure their gain during the treatment period as a measure of success. The use of the control group is a superior research design as we will review in future chapters, so you may consider the option of the control group if you have the capacity to serve those in the control group at a later time. You should be cautious, however, of selecting those for the experimental group on the basis of level of need and using the others as our comparison group because you have made the two groups unequal by your selection process. This makes the use of the comparison group questionable from the research standpoint.

Reamer (2010) offers more guidance on the above challenges related to risk, privacy, and informed consent. He admonishes social work researchers to be mindful of risks, sensitive to privacy, and diligent in obtaining informed consent. He offers many suggestions on how to pursue each of these issues.