Stories or Statistics?

Stories provide anecdotal evidence about the impacts of libraries while statistic provide evidence from scientific studies or operational research. Which are better in communicating the value of your library?

Nearly all stories and statistics deal with the direct benefits to patrons and not the indirect benefits to non-patrons. While researchers are working on the long term impacts and indirect effects, results from this are not available yet.

Table 1 gives the pros and cons of using stories rather than statistics. The bottom line is that both stories and statistics based on in-depth peer-reviewed research are important. It is not a case of one or the other.

Table 1: Pros and Cons of using Stories
Pros Cons
Stories paint a more concrete and subtle picture of the nature of the library service or experience and how it helped the patron. “You can prove nearly anything with an example.” This criticism of stories or anecdotal evidence is a valid one since your story could be the exception rather than the general experience.
Stories have more of an emotional punch than statistics, which makes them easier for everyone to understand. Stories can easily be biased by the experience of the teller. This might detract from their value when the issue is a highly controversial one.
Stories, especially when they involve real people with their real names and locations, are seen as more credible by many non-scientists who are unfamiliar with the technical issues behind many statistical results. Stories with real impact can have privacy issues that need to be handled carefully. If the names are changed to protect privacy, the stories lose a little bit of their credibility or require the reader to trust the story-teller.
Stories are basic observations, which often form the hypotheses that are later tested using more scientific methods. The basic concepts and connects described in a story cannot be refuted in the same way as a hypothesis can be in a scientific inquiry.
Stories are relatively easy to collect compared to developing statistics based on strong operational research or scientific hypothesis testing. Both operational research and hypothesis testing can be poorly done and lead to erroneous statistics. Unless the results are peer-reviewed in an anonymous fashion, the general public often has no way to separate the solid statistics from ones which are erroneous or biased.
The news media loves stories and seldom runs an article or report without leading with a story to illustrate the point. Yet, the news media can run with stories before they know whether in-depth research will back up these stories.