Confidentiality In Social Science Research
by Wes Jamison, WPI Interdisciplinary & Global Studies Division
Confidentiality can be defined as an explicit or implied guarantee by a researcher to a respondent in social science research whereby the respondent is confident that any information provided to the researcher cannot be attributed back to that respondent. Furthermore, the assurance of confidentiality carries with it the additional implication that non-researchers cannot discover the respondent's identity. Thus, confidentiality is an active attempt by researchers to remove any trace of respondent identities from the records.
There are two types of confidentiality. First, explicit confidentiality involves instances whereby the researcher provides verbal or written specification of the respondent's level of confidentiality. In this case, the researcher clearly defines and delineates her responsibilities to the respondent, and informs the respondent of the measures taken to assure the promised level of confidentiality. This type of confidentiality is openly negotiated between the researcher and respondent and may involve various levels of disclosure, which range from complete disclosure to absolute protection of identity. The essential issue in explicit confidentiality is that expectations and guarantees are clearly elaborated to the respondent before the research takes place, and those promises are honored throughout the research and publication of results.
The second type of confidentiality is more difficult to identify but nonetheless just as important. Implied confidentiality involves instances where the researcher implies through either word or deed that the respondent's identity and responses are protected. While the researcher may or may not be aware of the implied confidentiality, she nonetheless becomes responsible for upholding that implied agreement. Likewise, implied confidentiality also may involve an unspoken assumption by the respondent that her comments and replies will be un-attributable or "off the record." This is particularly problematic when dealing with people who have previously participated in social science research projects where confidentiality was assured. Often times, the respondent carries those expectations and agreements from the previous research into the current project, meanwhile failing to verbalize them to the current researcher. Hence, the respondent proceeds with the revelation of information under assumptions of confidentiality with which the researcher is neither familiar nor prepared to honor. This lack of expressed understanding is a common source of conflict between researcher and respondent. Therefore, it is extremely important in most contexts that researchers and respondents openly discuss and agree upon the levels of desired confidentiality, thus eliminating any incongruance between respondent and researcher expectations.
There are several practical issues involved with confidentially. First, there is the ethical responsibility to the people from whom information is gathered. This is because social science examines the public and private lives of people, including their ideas, beliefs, opinions, behaviors, emotions and attitudes. With this examination comes an extraordinary level of potential risk to the people being studied. A respondent in a survey or interview may experience social sanctions, peer and family scorn, social controls (e.g. police intimidation), termination of employment, or any number of other detrimental consequences if her views are disclosed. Similarly, individuals or corporations may lose competitive business advantages or have vital trade secrets revealed if the social scientist discloses the source of the information. Yet another potential risk comes in the publication of data that the respondent assumes to be confidential, thus causing embarrassment or legal penalty. For example, an IQP studying attitudes of students concerning sexual orientation carries enormous risk to respondents whom may either disclose their sexual orientation or oppose certain sexual orientations. The students who reveal their sexual orientation face potential scorn by peers if their identity and views are disclosed, while students who voice opposition to differing sexual orientations face threats of social control and sanction. Hence, since social science carries with it potentially enormous risks for respondents, social scientists are obligated to conduct their activities within a defined code of conduct which minimizes those risks. The most critical method for reducing the risks to respondents is the protection of the respondent's rights, privacy and welfare through an assurance of confidentiality.
Similarly, the assurance of confidentiality has practical benefits to the research project. With an explicit level of confidentiality, the respondent is much more likely to participate in a study. She is also much more likely to give honest and valid responses to questions. The deliberate obfuscation of data, or the production of misleading and dishonest answers, is a distinct possibility if the respondent perceives lack of confidentiality and therefore assumes an increased level of risk. Another common phenomenon is the reproduction of publicly acceptable answers. In other words, a respondent who perceives the threat of disclosure of her answers or identity is likely to provide data that conform to peer expectations and general public attitudes, therefore minimizing any associated risk. In the case of the IQP regarding student sexual orientation, a homosexual student who perceives a threat of disclosure is very likely to provide socially acceptable answers, (e.g. disapproval of homosexuality), in order to avoid social sanction and stigmatism. Hence, confidentiality both protects the respondent and assures the honesty and the validity of her responses.
A second issue involves an ethical responsibility toward other researchers. The boundaries of accepted method and ethics have been negotiated, tested and agreed upon over time by scholars with a personal stake in perpetuating their profession. Similarly, those methods and ethics have been refined through peer review and experience dating over the past two centuries. In the context of confidentiality, the accumulated body of knowledge and experience in the various social science disciplines verifies that confidentiality is an essential component of social science research. To assume a cavalier or lax attitude about that code of ethics violates the tenants of collegiality, mutual respect and accountability between academic disciplines that is critical to interdisciplinary, intellectual pursuits. More importantly, when a person conducts social science outside those boundaries, the researcher risks contaminating the research population for all subsequent researchers. For example, if a student IQP team does not adequately protect the respondent's identity in a survey concerning sexual orientation, and the respondent then suffers discrimination based upon revealed identity or responses, that person will most likely be reticent to participate in any future research. In other words, future students or faculty who desire to access that respondent for another social science IQP or project will be unable to gain the participant's confidence and cooperation because of the ethical breech committed by previous students and researchers. Furthermore, if the respondent belongs to an organization or community of like-minded individuals, the chances are very high that the organization will similarly be contaminated and less likely to participate in future research. In other words, "once burned, twice shy;" the fraternity, student group, or other organization whose member was victimized through a breach of confidentiality will most likely prohibit the future participation of its members in social science projects. Thus, social science projects that violate implied or explicit confidentiality not only injure respondents, they also penalize future researchers who seek access to the same individual or research frame.
A third issue that is increasing in importance is the legal liability that comes from conducting social science research. The courts have interpreted the social exchange between researcher and participant as a type of contract that carries with it certain legitimate expectations. Confidentiality is one of these expectations, and confidential participation should not carry intended or unintended penalties for the respondent. In other words, respondents in social science projects have come to expect that their participation should not harm them; if it does, they are increasingly likely to bring a SLAP lawsuit against the researcher or institution that supports the research. This issue is particularly acute in cases where businesses or interest groups perceive that the research results are detrimental to their interests. That said, the central tenant in assuring that participation in social science research does not penalize the respondent is confidentiality. This is particularly critical for the IQP because students are legally assumed representatives of WPI; hence, WPI becomes liable for their breech of confidentiality or for the detrimental impact of student social science projects upon the respondent. Although WPI and its students cannot entirely insulate themselves from lawsuits, following accepted ethical procedures within the various disciplines of social science provides protection from litigation. Furthermore, it cannot be overemphasized that a commonly accepted ethical obligation among the various social sciences is confidentiality. Thus, the practice of establishing and protecting confidentiality, or failing to do so, assumes legal ramifications.
In conclusion, there are myriad issues surrounding confidentiality, but suffice it to say that confidentiality is an essential part of social science. It provides an assurance of protection to participants, minimizes or eliminates their risk of participation, and involves the active effort of researchers to remove any trace of participant identity from the data. It protects colleagues and other professional social scientists, protects the profession of social science research, and partially insulates researchers and institutions from litigation. Consequently, students and faculty should assume that all responses are strictly and completely confidential unless stipulated otherwise, thereby taking active measures to eliminate any trace of respondent or subject identity from the results. Lastly, they should assume all participants in social science research, whether actively or passively engaged in providing information to the project, would enjoy and expect confidentiality unless clearly stipulated otherwise. Indeed, for ethical reasons, students and faculty should err on the side of caution in clarifying confidentiality with all the sources of their data. Maintaining this code of conduct is more difficult and time consuming, but to do otherwise is irresponsible and unethical. Bad social science hurts everybody involved, including the respondent, the researcher, the institution, and any future research projects. Perhaps the easiest way to cause harm to those parties is to violate the code of confidentiality that exists between a researcher and a respondent. Conversely, the easiest way to protect those parties is to follow established disciplinary guidelines that protect confidentiality and minimize risk.