





































In-Conference Workshops  313

Statistics That Matter
By Timothy D. Lincoln, Austin Presbyterian Theological Seminary

ABSTR ACT By disaggregating count data and creating questions 
about the quality of library services, theological librarians can create 
statistics that are useful for management, improvement of services, and 
demonstrating the value of the library in the educational enterprise.

This workshop addresses the need for librarians to collect and use 
statistics that matter. The life experiences that qualify me to lead 
this workshop include being a seminary library director and veteran 
writer of assessment reports along with (negatively) the fact that I 
was not a math major in college. Using librarian expertise and Excel, 
you can create statistics to help you manage your library better and 
communicate its value to constituents. In our time together, I will 
address three main areas: [1] sifting aggregate data to compute more 
useful granular-level statistics, [2] how to create home-grown ques-
tions about library quality, and [3] a cautionary reminder that data 
require interpretation to matter.

COMPUTING GRANULAR STATISTICS FROM AGGREGATE 
DATA

Librarians have a lot of data in hand that is counted more or less auto-
matically. Sometimes these data need further analysis to be helpful. 
Consider book circulation data. In aggregate, the number of circula-
tions for a given seminary library may be declining. A chart might 
look like this:



314  ATL A 2018 PROCEEDINGS

By and large, the chart shows a story of declining book usage. On 
an annual basis, circulations have fallen from approximately 18,000 
per year to fewer than 8,000. 

These same data tell a different story when you make some 
assumptions and do simple math. By defining core users as the total 
number of full-time faculty plus headcount total of enrolled students and 
doing division, these data are transformed into a computed statistic: 
the average number of book circulations per core user per year. The 
chart for this statistic looks like this:

SOURCE: Author. Actual circulation counts.

SOURCE: Author. Computed from circulation counts.

My Library: Historical Data

Average Circulations/Core User



In-Conference Workshops  315

The second chart also shows a decline in book circulations, but 
notice that the average hits bottom in 2013 (just under 20 circula-
tions per core user) and then goes up by ten circulations. Part of the 
reason that this second chart does not simply echo the first is that 
the number of students (the largest component of core users) has 
declined over time at my school rather dramatically. My point is this: 
that the computed statistic is more meaningful for you as a manager 
and for your faculty. 

An even more granular analysis of book circulation breaks down 
data by patron type. Librarians typically can get the gross numbers 
from the library management system. By using division, you might 
see a pattern of average circulations by user type that looks like this:

Average Circulations per Year by Patron Type

Once again, my point is that the computed statistic is more mean-
ingful than the gross number is. In this case, it may lead to a lively 
discussion about why DMin students use so many fewer books than 
MDiv students do, or perhaps a discussion about why faculty usage of 
books has dropped dramatically since 2011.

A second set of data that is easily obtainable is simple counts of foot 
traffic (gate counts). It’s one thing to know how many persons entered 
the library; it’s something else to determine how this relates to the 
core users of a library. In the case of my own library, total gate count 

SOURCE: Author. Hypothetical data.



316  ATL A 2018 PROCEEDINGS

has declined over time. By doing some division, however, a pattern is 
revealed about average visits per core user per year:

The computed statistic shows that the average number of visits 
per core user has not declined over time—despite the existence of the 
Internet and the advent of electronic reserves. So, in this case arith-
metic helps demonstrate that “library as place” continues to matter.

COLLECTING USEFUL DATA ABOUT LIBRARY QUALITY

Librarians often want to know about the quality of library services, not 
simply report activity counts or typical user behavior. The standard 
tool used in many theological schools, the Graduating Student Ques-
tionnaire, asks only two questions about library services:

 • Access to library services
 • Quality of library collections

Responses to these general questions don’t give you much to go 
on, especially if graduates are not satisfied. For instance, one would 
wonder which library services were not accessible, and why. Was the 
online catalog down? Were library hours inconvenient? To have more 
granular data about the adequacy of library collections, services, and 
spaces, you need to ask more detailed questions. As my library school 
professors used to say: This isn’t rocket science, it’s library science. 

SOURCE: Author. Computed from actual gate counts.

Average Visits per Core User



In-Conference Workshops  317

The table below suggests some questions suitable for faculty. Respon-
dents indicate their opinions using a Likkert scale (1: strongly disagree 
or strongly dissatisfied; 3: neutral; 5: strongly agree or strongly satis-
fied). It is a good practice to ask the same set of questions over time. 

Pertinent Questions to Ask Faculty

Does the library hold suitable books for 
use in the courses you teach?

Does the library hold suitable books for 
your research?

Does the library provide access to suitable 
full-text databases for use in the courses 
you teach?

Does the library provide access to suitable 
full-text databases for your research?

How satisfied are you with user training 
that the library provides to students?

How satisfied are you with the research 
assistance that staff provide you for your 
scholarship?

How satisfied are you with staff 
responsiveness to requests for book 
purchases?

How satisfied are you with support for the 
learning management system?

For several years, I have asked faculty questions like those above once 
a year at a faculty meeting. I distribute the questions on paper before a 
scheduled break. I literally stand in the doorway and collect completed 
questionnaires. In my context, this approach provides me with more 
data than other more sophisticated methods do (e.g., SurveyMonkey) 
and saves everyone’s time. I discuss findings with my staff. 

I routinely ask a different set of questions to students in a sampling 
of master’s-level courses and from students in all Doctor of Minis-
try classes. When it is time to write reports to accreditation bodies, 
I have in hand years of data to make the case that our library is good 
because its users say so. Behold, I show you a mystery: Collecting data 
is helpful!

CONCLUSION: PROPER INTERPRETATION

I conclude by reminding everyone that data need interpretation. 

This is not a sentence.
SOURCE OF highly unoriginal image: Author. With apologies to René Magritte.



318  ATL A 2018 PROCEEDINGS

In the science fiction classic The Hitchhiker’s Guide to the Galaxy 
by Douglas Adams, the universe’s most sophisticated computer 
concludes that the answer to the ultimate meaning of life, the universe, 
and everything is 42. One then needs to think long and hard about the 
meaning of 42. 

Collecting data like the examples used in this workshop (e.g., 
average number of visits to the library per core user) provides what 
lawyers still quaintly call “facts.” These facts need to be thought 
about in context. Given the way that professors at your school teach, 
it may be quite reasonable that typical book usage by students is 
low. In another context, low book usage would be a warning sign that 
the library isn’t collecting the titles pertinent to the curriculum.  By 
itself, one statistic, or even a ten-year-long set of data, is not self-
explanatory.

Having said that, however, I would rather go to the next budget 
meeting, library committee meeting, and my annual review armed 
with as many pertinent statistics as possible.


