Interviews with witnesses, victims, and family members of victims are the most common kind of information collection used by human rights organizations. For example, the Truth Commission for El Salvador completed approximately 7,500 interviews for their work, the Haitian National Commission for Truth and Justice some 5,500, and the Truth and Reconciliation Commission of South Africa hopes to do 75,000. With thousands of interviews, it is essential that the data be collected in an orderly, systematic way. This topic is discussed in considerably more detail in a later handbook in this series that covers interviewing theory and techniques for large-scale human rights data collection. The following discussion briefly introduces some of the key ideas in human rights interviewing for large-scale data collection.
Like all the kinds of information collection, interviews must be complete and comparable. This injunction means first that before beginning the questionnaire design, the organization has to define what they are going to need to know at the end of the project. Second, the organization must decide how they will manage the interviewers to ensure that each interviewer follows the same protocol.
Beyond comparability and completeness, there are several other ideas relevant to interviewing that might usefully be discussed here. The information that comes into an interview might itself be very complicated. The paragraphs below describe some of the complexities in this process.
A person (the complainant) comes to the organization to give information about a human rights event. The violent events being discussed may have happened only to the complainant. However, the complainant may be discussing events that happened to other people. Each person against whom one or more abuses were committed is a victim -- and the complainant may or may not be a victim.
Each victim may have suffered one or many abuses. The abuses may have happened at one time, or at several different times. Similarly, the abuses may have occurred at one place, or in several different places. Also, the abuses may have happened in the same time and place as abuses happening to other victims, or this victim may have been the only victim during some or all of the events. Each violent thing happening to a single victim is called an act.
Each act may have been committed by zero, one, or many identifiable perpetrators. The perpetrators may be identifiable as individuals (John Doe) or as organizations (the National Guard). Each identifiable perpetrator (whether individual or organization) may have committed one or several acts in the complainant's narrative. Each victim may have suffered at the hands of zero, one, or many identifiable perpetrators, and each perpetrator may have committed violence against one or several victims.
The structural complexity of a human rights interview can be staggering, which is why it is essential to organize the interview process carefully. If the process and the questionnaire are well-designed, the data will provide a stable start to the information management process leading to reliable reporting. The notes below provide some guidance to the steps necessary to a good interview design and technique.
188.8.131.52 Question design: After the organization has decided what information they need to collect, they must create questions that get that information. The questions should be un-ambiguous, and should mix closed-ended questions with open-ended questions.
184.108.40.206 Question ordering: In order to maintain the connections between the victims, the violations they suffered, and the perpetrators who committed the violations, it is very important to order the questions properly on the questionnaire. Often the need for absolute clarity about who did what to whom can lead to unavoidable circular or repetitive questions. Consider the following: "You said that John and Bill took you to the police station. Were you mistreated there? [yes, John beat me] How many times were you beaten? How were you beaten? Did others mistreat you as well?" The interviewer may ask several questions about each act in order to specify precisely what was done by whom in what sequence. This kind of detail and accuracy is vitally important for identifying perpetrators, accurately counting violations, and making legal claims. Representing this information properly on paper can be very difficult. There are many errors in questionnaire design analogous to the problems in database design described in Section 2.2.
220.127.116.11 Sensitivity to victims' dignity: Human rights organizations must always maintain as their first priority the restoration of the victims' dignity . The legislation that created the Truth and Reconciliation Commission of South Africa, for example, specifies in Article 11(a) that "Victims shall be treated with compassion and respect for their dignity." The human rights interview must provide the opportunity for the interviewee to begin a healing process by which, in the words of Agger and Jensen (1990), "private pain is transformed into public dignity." Therefore it is very important not to recreate the dynamic of a police interrogation. In an interview, the respondent answers questions that are asked by the interviewer. This dynamic tends to give the interviewer more power -- he or she controls the topics to be discussed and the flow and pace of the discussion. Furthermore, in the context of human rights violation research, the interviewer often has more education, more economic security, and more social capital than the respondent. In order to help the respondent maintain a sense of control and power, the interview process should be as conversational as possible, and the interviewer must give every cue that she or he is listening carefully and respectfully to the narrative. The goal of creating an interview that is conversational, respectful, and allows the respondent to retain control over part of the process may seem contradictory to the goal of getting all the necessary information. Good interview design combined with careful interviewer training can achieve both goals.
18.104.22.168 Cataloguing interviews: When a human rights organization begins to take hundreds or thousands of interviews, they must maintain careful control over what information comes in via the completed questionnaires. There are two aspects to this process: preventing sabotage and preventing loss. First, the staff managing the interviews must maintain a good numbering system of all the questionnaires. They need to have recorded unique numbers for each questionnaire so that they can identify it as an authentic questionnaire which was legitimately completed. By numbering the quesionnaires and recording who the interviewer was, as well as when and where the interview was completed, the organization can easily cross-reference this information with the information on the questionnaire itself in order to detect bogus questionnaires. With a systematically designed and unique set of numbers on the questionnaires, the organization can assure that questionnaires are not lost between steps of the process.
22.214.171.124 Training: The organization must provide extensive training of the interviewers and of the supervisors of the interviewers. Effective training of interviewers and supervisors is essential for assuring questionnaire control, comparability between interviews, and compassion in the interview process. Mock interviews are a good way to do this.
As I wrote above, all of these topics will be covered in much greater detail in the handbook about interviewing theory and techniques for large-scale human rights data collection.