Answering the Unanswerable
By Shirley Duglin Kennedy
As librarians and information specialists, we are relentless. Even if a question
seems impossible to answer, it goes against our nature to simply shrug our
shoulders and say, "I just can't find anything." So of course, the Computers
in Libraries session was packedfor a presentation by info guru Mary Ellen Bates
on "How to Answer the Questions You Can't Answer."
Unanswerable questions, Bates said, come in several varieties, each of which
demands its own strategy:
1. Questions no one knows the answer toExample: "How many
American flags have been sold
"Does no one really not know?" Bates asked. She suggested that researchers
start with relevant trade groupsin this case, the National Flag Foundation
or the National Retail Federation. In addition, she advised searchers to look
for anecdotal evidence in newspaper and magazine databases, while considering
whether there are "related indicators," such as an increase in military enlistments,
Hallmark's sale of more patriotic-themed cards, etc. "It's not as hard as it
seems to do some value-added distillation and extraction," Bates said.
And yes, go ahead and "throw it into Google," she said. You may come up with
some useful Web sites that, at the very least, will provide you with contact
information for someone who might be able to help.
2. Questions that require analysisExample: "Why is Wal-Mart
According to Bates, these are the most difficult types of questions for information
professionals because they require them to provide actual answers to customers
rather than just information. And yet, an analysisor an "executive summary"is "seen
as high value." So your focus should be on "what added value you can provide
to answer the question." Bates recommended doing a literature search; reading
SEC 10K reports, especially management's discussion and analysis, which tends
to be the most readable part of the document; and checking out analysts' reports.
Alas, Bates said, "Sometimes there is no one answer.... We can just
provide them with the perspectives that are out there."
3. Questions where no one really cares about the answerExample: "What's
the market for purchasers
of specialized carbon blades?"
It may seem trivial to you, but there's almost always some organization or
government agency out there that collects statistics that are related to it
in some way. Bates suggested checking such agencies as the Department of Commerce,
the Consumer ProductSafety Commission, NASA, the Department of Energy, and
trade associations.You can locate all of these organizations on the Web and,
even if the answer isn't right out there, you can find contact information
for someone who can help, such as a media representative or special librarian.
Another approach, said Bates, is to "look for parallel indicators." In this
instance, you might want to investigate purchase patterns for other carbide
products or other types of blades.
4. "Huh?" questionsExample: "I need examples of effective
For questions like this, Bates emphasized the importance of the traditional
reference interviewe.g., "What do you mean by HR departments? Within
our own organization? Within our industry? In our local area? In organizations
of similar size?" Then you need to find out "what we need to compare" to determine
effectivenesse.g., awards won, minimal employee turnover, training sessions
presented, etc. Finally, you have to determine what your customer wants in
the way of an answercase studies, journal articles, statistics/metrics,
lists of organizations that have won awards, etc.
5. Questions for which there is no chance of an answerExample: "How
often has our competitor's
network been hacked?"
Although someone, somewhere might very well have an answer to questions like
this, Bates said, the odds are good that "they're not going to tell you." You
may need to rely on what she termed "nontraditional sources of information." For
this example, you may want to start with network security organizations that
keep statistics for different industries or "white-hat hacker" Web sites, where
news of such might be posted. You may also want to try "black-hat hacker" sites,
where someone might have bragged online about the corporate networks he or
she has broken into.
Telephone research may also prove helpful. For example, a researcher could
ask how often a respondent thinks his or her competitor has been hacked. "Often
their answer reflects what their own reality is," Bates said.
Finally, she suggested taking a look at each competitor's 10K filings. "A
substantial hack may have been reported" because "if you lie to the SEC, sooner
or later you are going to go to jail."
6. Questions for which there are no sure answersExample: "Will
our industry be deregulated within the next year?"
"You're only going to find a soft answer" to questions like this, Bates said. "People's
best guesses." For questions on policy issues, she suggests checking with "people
[who] tend not to change from year to year," such as Capitol Hill staffers,
executive agencies' staff, legislative affairs staff in associations, and lobbyists.
7. Bum cite questionsExample: "I saw this article on The
New York Times' Web site. The title was 'Why Doctors Don't E-Mail.' Can
you get it for me?"
The key question to ask here, Bates said, is "If I can't find exactly that,
what would be second best?" This "elicits what, exactly, they're looking for," and
you might be able to turn up related articles that would be helpful.
Bates concluded with the following "tips, tricks, and sleights of hand" that
every researcher needs to know:
"Ninety-nine percent of the time, there is no one right answer.
Embrace uncertainty and ambiguity."
"Close enough is often good enough."
"Use peripheral vision. Look for something that looks close."
"Predictions are harder to find than past performance. Offer
to extrapolatee.g., 'If the future continues like the past, it may
go like this.'"
"Phone research often unearths valuable, soft information." If
you're not good at thisand many of us
are notyou can hire professionals to do it for you.
Offer analysis or an executive summary. "Less is more."
"Packaging matters," so you want to make your client report "as
un-dense as possible." Remember that librarians are more comfortable "in
the sea of information" than end users. Use "lots and lots of white space."
"Sometimes, the question that they ask is more negotiable than
it appears.... You can turn an unanswerable question into an answerable question."
"People are much more willing to give you information than not...." Or
they can refer you to someone else who can be helpful.
"The thing that we can do is provide answers. Not just information....
That's where we show that we're irreplaceable."
Shirl Kennedy is an Information Today columnist and
the electronic resources librarian at the Walker Management Library at Vanderbilt
University's Owen Graduate School of Management. Her e-mail address is email@example.com.