Performance measurement in libraries and information services

1
Performance measurement in libraries and
information services

• Part II. Library statistics Background
• Prepared by Jurgita Rudzioniene,
• Training Centre for LIS Specialists in Lithuania
• City University, London
• 2001 July

2
It is remarkable that statistics is both one of
most useful and powerful tools in data analysis
and also one of the most feared and hated
subjects of study Vaughan L., 2001
3
"I like to think of statistics as the science of
learning from dataIt presents exciting
opportunities for thosewho work as professional
statisticians. Statistics isessential for the
proper running of government, central to
decision making in industry, and a core component
of modern educational curricula at all levels."
Jon Kettenring ASA
President, 1997
4

• Definitions

American Heritage Dictionary defines statistics
as "The mathematics of the collection,
organization, and interpretation of numerical
data, especially the analysis of population
characteristics by inference from sampling."
The Merriam-Websters Collegiate Dictionary
definition is "A branch of mathematics dealing
with the collection, analysis, interpretation,
and presentation of masses of numerical data."
5

• The steps
• of statistical analysis involve
• collecting information
• evaluating it
• and drawing conclusions

6

• PUBLIC AWARENESS (PA)
• Library data are useful information about past
  and present activities in order to prepare for
  future ones. Reports are useful for several
  categories of people
• the head librarian will make management decisions
  based on these reports
• library staff should also have a global view on
  all library activities and must understand why
  they have to engage in the daily collection of
  data
• informing library users about the librarys
  activities makes them aware of the reasons why
  certain rules and limitations are effective and
  why certain services are not free of charge
• making reports public to library patrons and
  staff is a good way to give them a sense of
  involvement and so solicit new initiatives and
  constructive criticism
• financing bodies must regularly informed. This is
  best done by annual reports (calendar year,
  academic year or fiscal year)
• is the main official document describing and
  defending the librarys mission and how its aims
  were reached throughout the past year

Annual report
7
PUBLIC RELATIONS (PR)
Distributing informative brochures amongst
patrons
Contacts with the (local) press
Also electronically on the internet
Combining PA and PR aspects
8
Towards understanding and utilizing statistical
methods

• background
• types of statistical data
• graphs and descriptive statistics
• basic concepts of inferential statistics
• sampling methods
• statistical tests
• advanced statistical methods

9
Recognizing the types of data

• 1. Nominal data
• 2. Ordinal data
• 3. Interval data
• 4. Ratio data
• Data conversion

10
1. Nominal data (scale)

• Code data (arbitrary and symbolic)
• Limited calculations (the average between male
  and female?)
• Frequency calculation possibilities

• Nominal data are obtained when numbers are used
  to arbitrarily label categories. The only
  calculation that can be applied to nominal data
  is to note the frequency of occurrence of each
  category

11
2. Ordinal data (scale)

• Ordinal data contain information as to better or
  worse, or greater or less, but they do not tell
  us details as to how much better or how much
  greater.

• Code the answers to the questions
• Coding is not completely arbitrary and must
  follow an order from high to low and vs vrs

12
3. Interval data (scale)

• There is no absolute zero point
• Bonus point system

• Interval data provide not only
  greater-than-or-less-than information, but also
  details on how much greater than or less than.
  However, interval data have no absolute zero
  point, so that we CANNOT use comparisons such as
  twice as many or half as much with interval
  data

13
4. Ratio data (scale)

• Ratio data provide not only greater-than-or-less-t
  han information, but also details on how much
  greater than or less than. In addition, there is
  an absolute and non-arbitrary zero point so that
  we CAN use comparisons such us twice as many.

• Measurement of whether a given response is
  greater than or less than other responses
• There is an absolute zero point

14

• Data conversion

NO
YES
Ratio data
4
4
Interval data
3
3
Ordinal data
2
2
Nominal data
1
1
15

• References
• Busha Ch. H., Haster St. P. Research methods in
  librarianship techniques and interpretation.
  N.York,
• London etc. Academic Press, 1980. 417 p.
• Egghe L., Rousseau R. Elementary statistics for
  effective library and information service
  management.
• London Aslib, 2001. 120 p.
• Slater M. Research methods in library and
  information studies. London The Library
  Association,
• 1990. 182 p.
• Line M. B. Library surveys an introduction to
  the use, planning procedure and presentation of
  surveys.
• 2nd ed. London Clive Bingley, 1982. 162 p.
• Vaughan L. Statistical methods for the
  information professional a practical, painless
  approach to
• understanding , using, and interpreting
  statistics. Medford, N.J. Information Today,
  Inc., 2001. 209 p.
• (ASIST Monograph Series)
• The American Statistical Association.
  http/www./library.city.ac.uk/search/sJournal/