STORIES FOR RESEARCH 323norms and beliefs, the distinctions between fact and fiction are meaningless: Inventions and em-
bellishments may be more revealing about beliefs and judgments than responses to carefully
written questionnaires. Though we advise narrative researchers not to worry about the distinction,
we must be clear that we are not collecting, presenting, and interpreting narratives as if they were
accurate accounts of events. We treat them, rather, as revealing accounts of norms and beliefs.
Analyzable and Reanalyzable Stories. If the purpose of the research is to present and publish
observations that are compelling to others, then good stories for research are in a form that can be
examined, analyzed, and archived for reexamination by other research teams. Stories for research,
in contrast to stories collected for other reasons, need to contribute to interpretive arguments that
can withstand the rough-and-tumble of scholarly debate. Interpretive researchers must acknowl-
edge that there are no definitive interpretations.
To contribute to scientific discourse, we must demonstrate how our views are grounded in data
and encourage others to challenge our views based on the same data. This requires the careful,
time-consuming, and often costly reproduction of transcripts, notes, and other field data and, with
proper safeguards for confidentiality, the open sharing of field data. It is not possible to reproduce
and share every note and impression, but the credibility of our interpretations are only strength-
ened by allowing others to examine our sources.^4
Many stories relevant for social research are oral, rather than written, and are told and heard,
rather than read. Such storytelling is a performance, with each performance varying depending
on the setting and listeners. In contrast to written storytelling, oral storytelling is elusive and
evanescent, making analysis all the more difficult. (Advances in storing and examining digital
audio and video, however, may facilitate the reproduction and analysis of oral storytelling events.)
At present and in part because social researchers communicate findings through publication, oral
renditions are most often translated into written texts prior to analysis. As with all translations,
some things are lost, but written stories have the advantages of providing fixed, readily available
texts that can provide the foundation for evidence-based interpretations. Written texts can be
reproduced and included in publications to document and justify interpretations to enhance the
credibility of the research.
It is crucial that the written text produced by the researcher reflect the story that the storyteller
wants to render. To make sure that we captured their stories, we took the transcriptions back to the
storytellers at least once and sometimes several times. For our research purposes, a story was not
treated as a final text until the storyteller was fully satisfied that the written story was the account
he or she wanted to convey.
Adequate Numbers. Although much can be learned from the careful reading of a single story,
interpretations and analyses of the sort we were engaged in are strengthened and broadened when
based on an adequate number of diverse stories. Themes and patterns are rarely developed in a
single story, nor can we be confident that the insights contained in a handful of stories will enable
the rendering of interpretations of significant breadth to have meaning beyond the single setting
or individual experience. Selecting units of analysis remains a central question in research design.
Interpretive researchers who closely examine one or a few life histories or cases are unlikely to
rely on a single story from each individual, drawing, instead, on the collage of narratives that
constitute an individual life or a case. We designed our research to look for patterns of meaning
across organizational setting and form of street-level work.
Although there is no formula for what constitutes an “adequate” number of stories, the guide-
lines are similar to those for intensive field interviews. Researchers should collect and examine