Skip to Main Content

Oral History - Methodologies and Sources

This guide brings together information on oral history methodology as well as a list of select oral history collections and resources at Duke Libraries.

Interviews, even tightly controlled ones, have the potential to wander.  A quickly tiring narrator, the vagaries of memory, the tendency to pursue tangents – these can derail good interviews more often than add to them.  Good research and questions will prevent if not entirely eliminate these obstacles to obtaining a research-quality interview.

What constitutes good research in terms of preparing for an oral history interview is relative to the narrator and the topic in question.  It is possible that so little is known about the narrator or topic that only broad contextual research is necessary or even possible – although there is a point where you may want to reconsider why the subject is being interviewed – and that questions are really more like prompts.  At the other side of the spectrum is the interview where critical gaps exist in an otherwise well-known story, and the interview is less about discovery than answering a narrow band of questions.  Most interview situations lie somewhere between these extremes, and in general research should result in a kind of cultural fluency -- familiarity with the political, artistic, socioeconomic worlds inhabited by the subject -- that can allow for a meaningful discussion between interviewer and narrator.  And, of course, the definition of the project should drive the direction of the research.  In addition to performing basic research using secondary sources, it is useful also to consult WorldCat and Google – a narrator may have already done an oral history, a not uncommon occurrence, answering research questions you have or creating new ones.  Then there’s going straight to the source:  in developing an oral history, a pre-interview meeting with the narrator can illuminate topics, and, importantly, create a sense of partnership.  Successful documentarians will often invest significant time (days, weeks, months, years) with their subjects before taking out their cameras and their microphones, ensuring that a relationship precedes what is recorded.  Although oral historians don't always have this luxury, building a rapport with the interviewee, even if it's only in the few minutes before an interview, will lend itself to a comfortable session and a more complete history given.

Building questions, like research, depends on the nature of the interview.  If a conversational interview is anticipated, an outline that guides the discussion will maintain the relaxed feel, while a written set of questions could be helpful in a formal setting, or in a situation where the participants feel more comfortable having an explicit structure.  In either case:

  • Construct open-ended questions that allow for elaboration by the narrator.
  • Keep questions simple.  Multi-part questions are rarely answered completely.
  • Many questions can function as a springboard to unanticipated topics – be prepared to modify questions brought to an interview.
  • While every interview is different, each should contain questions establishing the narrator’s identity: full name, date and place of birth, and depending on the direction of the interview, selected genealogical information.  These questions are usually best addressed at the opening of the interview, and, as standard questions, can function as a warm up, particularly preceding questions regarding early life experiences.
  • As a starting point for constructing general questions, check out Story Corps's Great Questions page.  It's a particularly good resource for family-and-friends oral histories.


Do oral histories conducted as part of research projects in academic or research institutions require the support of an institutional review board (IRB)? Well, it depends...

For years there has been debate regarding IRB oversight of oral histories, and IRBs at different institutions set different policies and often work on a case-by-case basis because, as we can imagine, oral histories can run the gamut of research topics and uses.  In general, though, IRB oversight of oral history has been considered an administrative hurdle having the potential to impede humanities research, under which the bulk of oral histories fall.  In early 2017 the federal government revised its Policy for the Protection of Human Subjects, to clarify that oral history need not be considered research in the spirit that the policy was originally developed in 1991, and therefore did not require IRB approval.  The revised policy is scheduled to go into effect January 19, 2018 (more here in the Journal of Higher Education), although it is possible that the date will be pushed back to allow further evaluation.

Even with the new federal policy, research institutions (like Duke) can implement their own IRB policies, to help protect human subjects, ensure ethical practices, and create valid research.  These policies may encompass oral history.  The best course of action when creating oral histories within a research institution setting is to contact the IRB office.  You can contact the Duke IRB office here: