Data Collection on Sensitive Topics
“Research participants will discuss very sensitive subjects if given guarantees about how the information will be used and if assured that their anonymity will be protected.”
Research participants often do reveal highly personal aspects of their lives to researchers, beyond what people would normally disclose even to close friends. While it is not possible to state whether a one-off encounter is better or worse than multiple-encounter research in collection such information, it is important to note that the quality of the data collected may be dependent on the researcher’s ability to develop both an intimate and disinterested relationship with the participant.
There are many definitions of sensitive research, ranging from those that refer to the topic of investigation to those that encompass the whole of the research activity.
Some obvious examples of research AFS are asked to undertake are:
Victims of Crime Illegal Habits Underage Behaviours
Business (Information) Grief/Abuse Health Treatment
Respondents feel Threatened
In approaching these topics interviewers must understand the level of threat respondents may feel in their participation. Researchers and supervisors will identify and inform interviewers of these possible threats and how to navigate them. The first area is an intrusive threat which deals with areas that are ‘private, stressful or sacred’ such as sexual or religious practices. The second is a threat of sanction, which relates to studies of deviance and involves the possibility that research may reveal information that incriminating in some way – take fraud as an example. The third type of threat that may be imposed by sensitive research is a political threat and in these situations researchers may trespass into areas that involve some sort of social conflict or social perception which may force the respondent into a lie 'please state your current income' is an area where this occurs, especially if their role definition, age or affinity class is in discord with the actual reality. Political threats also occur when there is some perceived interest in lying, “yes we need more government investment”, “yes I pay too much tax.” This is an important consideration for researchers studying sensitive topics who do not wish to exploit the community, for example, migrants, ethnic minority groups or low-income groups.
Approaches to Collecting Data
Researchers of sensitive topics can choose a qualitative design using the in-depth interview as their preferred method of data collection. There is a range of issues that arise when qualitative interviewing, one of the main issues raised is that these interviews are often stressful for both the researcher and the interviewee.
Researchers need to be able to prepare themselves to physically and emotionally disengage at the end of the research.
Interviewing quantitatively over the phone has become a more attractive way to provide an arm's length, to better navigate sensitive topics and openly talk on case specific items – this can disarm many of the potential hazards and sensitivities of running such projects in person.
In most cases, the added advantage is interviewing is best achieved when the relationship of interviewer and interviewee is non-hierarchical, like it can be over the phone.
Reactions to Sensitivities
Collecting data in sensitive research can sometimes be a very difficult task and is often fraught with problems. AFS have developed a number of methodological innovations which has facilitated the collection of data on sensitive topics. These learnings have focused on, the asking of sensitive questions on surveys and methods to enhance the protection of confidentiality, as research participants are often willing to discuss very sensitive subjects if given guarantees about how the information will be used and if assured that their anonymity will be protected. AFS ensures that any potential questions on the authority of the study can be immediately provided to the respondent, items such as sponsor representative contacts and where immediate professional advice can be sought if required.
Protocols: Setting Up for the Introduction of Sensitive Topics
The first AFS protocol is developed for use in telephone screening to identify individuals at high risk for adverse emotional reactions – testing their willingness and depth of reaction early in the introduction. This is not to bias research from people in certain circumstances but rather to read the potential respondents temperament to handle being researched at this time. The second protocol guides interviewers' responses to emotional distress expressed by participants during in-depth research interviews and suggests follow-up questions or ways to handle anxiety that might be expressed during the interview – and redirect the respondent to the survey.
Part of collecting such data involves researchers taking their participant’s emotions into account. This emotion can take the form of verbal and non-verbal communication. For example pauses, silences and non-verbal emotional displays such as tears and embarrassment also need to be measured in the instrument (one practice is at the completion of the interview, the interviewer indicates the emotion displayed by the respondent.)
A protocol which has worked well is if the instrument begins with generalised associations such as 'people in this situation have communicated these outcomes/behaviours/feelings/emotions' to edge the respondent to open up. The hurdle here is not priming the respondent - so the more general the association the better. These questions are a simple way to gauge respondent receptivity and are not necessarily used in analysing results.
Implications of researching a Sensitive Topic
Researchers often ask their own questions about accessing vulnerable and dis-empowered people for research purposes. Firstly, the benefits of undertaking the research have to outweigh the risks of undertaking the research with certain respondents - this is a difficult problem as some of the issues surrounding sensitive research are not always apparent at the outset and reactions not easy to test without some piloting. Researchers often cannot predict how they, interviewers or the participants will be affected, as they often do not know in advance what may come out of the research. The best advice is to seek people who have implemented parallel surveys in the past, question the approach taken,the curve balls which resulted and how they handled it.
Back to Best Practices Library
