Fopl ola sc stats with molyneux

Overview of FOPL's Ontario
Public Library Statistics and
Measurements Report
Dr. Robert Molyneux, MLIS, PhD
Stephen Abram, MLA , FOPL Executive Director
OLA Super Conference
Jan. 29, 2015

Introductions
Today’s Speakers
Dr. Robert Molyneux, MLIS, PhD
Stephen Abram, MLA , FOPL Executive
Director
2

FOPL Webinars in the Stats Series
August 14th
, 2015 Overview of FOPL's Ontario Public Library
Statistics and Measurements Report
 Speaker: Robert Molyneux, MSLS, PhD
August 17th Overview of FOPL's Market Probe Canada Public
Opinion Poll of Ontarians and Public Libraries
 Speaker: Carol French, Senior Vice President, Research &
Client Services, Market Probe Canada
August 28th Strategic Use and Insights from FOPL's Ontario
Public Library Statistics, Polls, and Measurements
 Speaker: Stephen Abram, MLS, executive director FOPL
3

Context of FOPL Stats and
Measurements
 Statistics and Measurements Strategies
– Participate in CLA task force on national statistics
– Host 2 iSchool symposia on measurements for libraries
– Lobby for open data for public libraries
– Publish analysis of Ministry data collection for 2001-2013.
– Publish Market Probe opinion polls for 2015 (building on
2001, 2006, 2010 polls)
– Host and record webinars
– Coming Soon: FOPL Index of Community Engagement
 Thank you to the Ministry of Tourism Culture and Sport for
support
4

Limits of Library Statistics
 Library statistics are very complicated and complex.
 Most libraries do not have a strong culture of
measurement.
 What is the difference between statistics, measurements,
polls, etc.
 All numbers have inherent risk when communicated and
interpreted.
 Data - Information - Knowledge - Action/Decisions
5

Potential Comparisons
 Compare by size of population (or any other
data point (expenditures per capita, etc.)
 North – South (e.g. SOLS vs. OLS North Libraries)
 Rural, Remote, Town, County, Suburban, Urban
Libraries
 Special Groups – Francophone, First Nations, etc.
 Regional comparisons (e.g. libraries around Ajax or
libraries around Waterloo)
 Handpick a peer cohort
6

What do we normally use our data
for?
 Strategic Planning
 Program Analysis
 Budget justifications and comparisons
 Tracking success, growth, decline
 Decision support
 Finding libraries like ours to compare our experience to
theirs
– Which means they can be a directory to libraries like
mine for benchmarking and cohort analysis
– Comparing like with like is important
 And more . . .
 We will add trying to get a sense of the health of and trends
in Ontario’s public libraries
7

If you’re not collecting the
data to inform an action,
then why are you
expending the effort?
Making Decisions and Choices

Library data have a long history
9

Library data have a long history
 We have fragmentary numbers of collections
of a number of these libraries
– And like modern library numbers, we are not
always sure exactly what they mean
 Adriano Balbi, A Statistical Essay of the
Libraries of Vienna and the World [1835]
– First modern attempt at comparing libraries in
major European cities using published statistics
about them10

Balbi’s observations
 “disheartened by a disparity of opinion…”
 “only approximate data”
 “exaggerated” numbers in pursuit of prestige
Then a wonderful discussion of the problems of
comparative library data
11

Types of library data
 Balbi was dealing with reports by visitors to
various libraries at different times who
recorded estimates they heard from a variety
of people
– One-time studies done by different methods
 Episodic surveys
– Attitudinal surveys—particularly users and
non-users
– Data collection on fugitive or new subjects
12

 The data we are going to discuss are
systematically collected, annual data,
professionally compiled from surveys of
Ontario public libraries by the Ministry of
Tourism, Culture, and Sport
 Available from 1999-2013 in pdf
 2014 Coming Soon!
13

Now converted to csv files
 The Ministry has reissued these data in csv
(“comma-separated values”) which means
they can be read into a spreadsheet program
such as Excel or LibreOffice Calc readily.
 In other words, there is not a chance of
introducing error when you rekey data.
 This is a tremendous boon to studying our
libraries using these data.
17

What can we do with these data?
 We can look at any variables we choose for
individual libraries in one year or all libraries
in one year.
– For example: How big were the budgets of
Ontario libraries in 2013?
 With a good bit of work, we can rearrange
the data and look at the select variables
through time—that is, trends.
– Say: What happened to their budgets from 2001-
2013?22

We did a bit of both and more
 The report is a sampler of what can be done
with these kinds of data with the focus on a
province-wide view, not individual libraries
23

We did not use all data from this series for our
report
 Small number of variables
 In order to analyze trends properly, we only used
data from libraries which reported each year. For this
study, that number is 301 libraries
 Years 2001-2013
 We separated them into 9 “Bands”—8 by size plus
the First Nations’ Libraries in a 9th
Band.
– The Ministry did the same thing
– This is common practice in this kind of analysis
24

LET’S TAKE A TOUR
THROUGH THE REPORT
25

First the Primer
 The big story is the consistent characteristic of the
library world that affects about everything:
– Skewed distribution: a few large libraries and many
small ones
 In 2013, the 10 largest libraries (of 300+) had 60% of the
total circulations and 54% of the total expenditures.
 We must take these characteristics into account in
analyzing data
– Hence, our size “Bands” which follow Ministry practice
26

The Primer, then, informs the analysis in the FOPL
Reports
 Given it is a sampler
– We segment by size of library in “Bands”
 Same as those used by the Ministry with a difference: First Nations’
Libraries are analyzed separately in those tables where we use Bands
– Another common tool is the “Rank Order Table”
 Sort libraries in order by their reported data. That, is rank their results
by the reported data or statistics calculated from these data
– Most commonly per capitas. Dividing, say, circulations, by the
resident population served by the library
 We combine this technique with analysis of Bands.
– Our focus, primarily, is the state of the provinces libraries and trends
affecting them
27

There are many other things you
could do with these data
 This is a rich series
28

2013 data per capita and per
cardholder
 Thirteen ratios, all libraries
 The ratios are largely those we focus on in
the rest of the report.
 As the Primer showed, per capitas allow
apples to apples comparisons of libraries of
vastly different sizes
– You may be small, but you may be doing a better
job with what you have than bigger libraries.
30

Using the Spreadsheet Versions
 You have pre-crafted tables in the report(s)
 You could take the spreadsheet and mosey
around in it a bit. Sort by this or that—it is a
very busy table and one hard to show in
slides
 FOPL can make the spreadsheet available to
the members on request.
 We are also happy to do custom analysis for
you on request for a quoted fee.
31

A bit of caution if you work with
spreadsheets
 Save a copy of the original spreadsheet
 Did I mention saving a copy of the original
and don’t change it
 Make another copy for analysis. If you make
a mistake, you always can go back to the
safe copy
 Working with spreadsheets requires caution
– You think you did something but you can’t audit
what you have done
32

Annual Population and
Circulation, 2001-2013
 Trend analysis is a bit different
 Of all libraries which reported in any year,
301 reported in each year
 These tables are complex
– We will see them again, so let’s take a look
33

Traditional library measures are
steady
 OTOH: New things are growing
36

Rank Order Tables
 Circulation per capita and per active
cardholder, 2013, by Bands
44

Rank Order Tables
 Expenditures per capita and per active
cardholders
48

Conclusions
 This is the beginning. A first shot based on
best guesses of where to look.
 There are other ways of studying libraries
such as qualitative surveys of a library’s
users and their non-users.
– Given the rapidly changing information
environment in libraries, quicker surveys likely will
be a part of the future of data gathering to support
decision making.
56

Next Step
 Develop a NEW FOPL Index of Community
Engagement for testing and discussion.
 Can we combine in various ratios the hard
copy and digital transactions and attendance
of our library members to compare libraries
on a more fulsome basis than ‘circulation’?
57

Questions?
www.fopl.ca
Stephen Abram,
Executive Director
416-395-0746
sabram@fopl.ca

Thank You
www.fopl.ca
Stephen Abram,
Executive Director
416-395-0746
sabram@fopl.ca

Fopl ola sc stats with molyneux

More Related Content

What's hot (20)

Similar to Fopl ola sc stats with molyneux (19)

More from Stephen Abram (20)

Recently uploaded (20)

Fopl ola sc stats with molyneux

Editor's Notes