User behaviour in largescale resource discovery contexts

1
User behaviour in large-scale resource
discovery contexts

• Dick Hartley and Helen Booth
• CERLIM
• Manchester Metropolitan University

2
User behaviour introduction

• Aims
• Methods
• Results
• Comments

3
User behaviour aims

• CERLIMs role was to study user behaviour in
  the searching of union catalogues whether
  physical or virtual

4
User behaviour methods

• Data collection consisted of three distinct
  phases
• End-user searches
• Interviews
• Focus groups
• Data collection took place in three different
  university libraries LSE, Leeds and Strathclyde

5
User behaviour end-user searching

• Manufactured but realistic queries developed with
  the advice of staff from clumps and COPAC, 4
  types
• Find a common item in a library close by
• Lot of information provided (what used?)
• Find an obscure or unique item
• Subject query giving large results

6
User behaviour example queries

• You urgently need to get hold of a copy of
  Clinical Medicine edited by Parveen Kumar and
  Michael Clark. You need to get the most up to
  date edition you can find in the nearest library.
• You are doing some research into monasticism and
  are interested in books on monks behaving badly.
  Someone has told you about a book that was
  published some time in the late 90s. They arent
  sure of the title but think that the author might
  be Justice or Jestice.

7
User behaviour searching

• Each searcher undertook 4 queries, 2 on COPAC and
  2 on local clump
• No training but exploration time
• Stop when
• searcher was satisfied that the appropriate
  search result had been achieved,
• searcher was not satisfied but did not want to
  proceed further,
• searcher was not able to proceed further,
• searcher was fed up and wanted to go to the next
  task.
• Searches logged

8
User behaviour interviews

• Post search interviews explored
• Search options used
• Selecting libraries
• Results
• Error messages and feedback
• Problems encountered
• Search session
• Features liked, disliked and desired
• Overall feelings

9
User behaviour focus groups

• Took the view that librarians are users of union
  catalogues
• Given their knowledge and experience, used focus
  groups to enable discussion and ideas to flow
• Discussion based on a checklist of topics
• Taped and notes

10
Data analysis

• Search logs analysed into a large spreadsheet
  with a search as a row and a feature used as a
  column
• Enabled easy analysis of features used
• Interviews and focus groups transcribed and
  manually analysed.

11
Results

• Profile of the end users
• Briefly consider results of searches
• Concentrate on attitudes and perceptions of end
  users
• Briefly pick up some points from focus groups of
  library staff

12
Profile of end users (1)

• Mix of genders
• 20 female and 14 male
• Spread of ages
• 21-30 19
• 31-40 04
• 41-50 05
• 51-60 05
• 61-70 01

13
Profile of end users (2)

• Range of academic levels, FT and PT
• 
  FT PT
• PGCE 1
• Taught Masters 4 3
• Research student 9 2
• Research staff 8 2
• Academic staff 3 1
• Admin staff 1

14
Profile of end users (3)

• Wide range of disciplines represented
• Accounting and finance, Archaeology, Biology,
  Counselling, Economics, Education, English,
  Environmental science, Genetics, Information
  management, Linguistics, Marketing, Media and
  communications, Medicine, Neuroscience,
  Philosophy, Politics, Psychology, Sociology,
  Statistics

15
Searchers use of e-resources

• Internet search engines
• Daily 28
• 4-5 times per week 1
• Once or twice per week 5

16
Searchers use of e-resources

• Bibliographic databases
• 3-4 times per week 3
• Once or twice a week 13
• Every other week 6
• Once or twice a month 5
• Less than once a month 6
• Have not used yet 1

17
Searchers use of e-resources

• OPACs
• Daily 3
• 3-4 times per week 3
• Once or twice per week 15
• Every other week 6
• Once or twice a month 4
• Less than once or twice a month 2
• Never 1

18
Results from searches

• 30 out of 228 searches were abandoned without
  getting results
• Numerous spelling errors, some of which were
  never noticed and some of which produced results
  egs
• Econimics, Bryon, Liguistics, Parlement,
  Shapespeare

19
Results from searches frequently used features

• CAIRNS
• Refine search
• Select libraries
• COPAC
• Help
• LanguageEnglish
• Select libraries
• InforM25
• Map
• List of libraries

20
Results from searches

• Searches for Books
• Almost equal use of author, title and isbn as
  search criteria
• Searches for periodicals
• ISSN followed by title most used search criteria
• But different searchers used a wide range of
  search criteria both singly and in combination

21
End-user perceptions of Union Catalogues(1)

• The term was almost completely unknown even by
  those who were using them
• Two or three users hazarded a guess that they
  were associated with Trades Unions.

22
End-user perceptions of Union Catalogues(2)

• End-user expectations of the performance of a
  union catalogue are heavily influenced by the Web
• Predictably this referred largely to Google but
  also to Amazon
• If cannot learn to use a search tool very rapidly
  then users would go to Google or Amazon
• Some demand for greater information about book
  content

23
End-user perceptions of Union Catalogues(3)

• End users expect search tools to be easy to
  master
• A typical comment was If I cant get the hang of
  what I am doing in the first half an hour then
  Id reject the package and look for something
  else.

24
End-user perceptions of Union Catalogues(4)

• Whilst there was the odd exception, most
  searchers were unwilling to wade through a large
  number of retrieved items
• Various approaches to dealing with large number
  of hits
• Narrow search
• Find out more about topic then try again
• Give up and use the Web instead

25
End-user perceptions of Union Catalogues(5)

• End users have problems with on-screen language
  that is taken for granted by information
  professionals.
• Examples quoted to us included
• Miniclumps
• Holdings
• Tag
• Z39.50
• Anything (COPAC search option)

26
End-user perceptions of Union Catalogues(6)

• End users do not like the presence of duplicate
  records in search output
• (fortunately for them they do not have to develop
  matching algorithms to remove dupes and only
  dupes!)
• ( how do they react to duplicate retrieval of the
  same website by Google?)

27
End-user perceptions of Union Catalogues(7)

• Many users expect Google-like speed of
  response.
• This was evidenced by the number of abandoned
  searches.

28
Librarians perceptions of end-users and union
catalogues

• End-user awareness
• I would say that I dont think students are
  particularly aware but some researchers might be
• Researchers use
• Location
• Bibliographic details
• Researchers willingness to travel

29
Librarians and union catalogues

• Well aware of the concept
• Quoted a far wider range at us than those which
  were the focus of our study eg RLIN, WorldCat,
  even BUCOP!
• Limited trust in search output
• Uptodateness of physical union catalogues
• Doubts about efficacy of Z39.50 searches for
  virtual union catalogues

30
Librarians views on scope of union catalogues

• Serials union catalogue is the most important
• After serials concentrate on rare books
• Some support for regional union catalogues
• No enthusiasm for subject based union catalogues

31
Facilities wanted in union catalogues

• Ability to sort by date
• Ability to sort by author, title
• Ability to place own library first
• Ability to sort geographically
• Ability to link search output to inter-library
  loan software
• Ability to pass search output to reference
  management software

32
Comments and observations (1)

• COPAC usage data shows that there is clearly a
  demand for union catalogues
• How much greater could that demand be if there
  was greater awareness amongst potential users of
  the existence of such tools?

33
Comments and observations (2)

• There is a need for a major awareness-raising
  effort to increase use of union catalogues and
  other tools into which there has been
  considerable investment

34
Comments and observations (3)

• Designers need to take into account user
  expectations
• Simple interface
• Speedy response
• Manage expectations eg indications of processing

35
Comments and observations (4)

• Investment in SUNCAT appears to be the correct
  decision
• Is it worth investigating the notion of what is a
  rare book and concentrating on adding those to
  union catalogues of books?

36
Comments and observations (5)

• Perhaps useful to recall Mooers Law
• an information retrieval system will tend not to
  be used whenever it is more painful and
  troublesome for a customer to have information
  than for him not to have it

37
The full report is

• Helen Booth and R.J. Hartley
• User behaviour in the searching of union
  catalogues an investigation for Work Package C
  of CC-interop. Manchester CERLIM, February 2004.
• (http//ccinterop.cdlr.strath.ac.uk/documents/f
  inalreportWPC.pdf )

38

• Thank you to all colleagues at CAIRNS, COPAC and
  InfoM25 and Tracey Stanley at University of Leeds
  for help in organising the searching, interviews
  and focus groups and devising the questions

39

• Thank you for listening!

40
(No Transcript)