ProSeBiCA Development of New Library Services by Means of Conjoint Analysis

1
ProSeBiCADevelopment of New Library Servicesby
Means of Conjoint Analysis

• Antonia Hermelbracht, Bettina Koeper
• Bielefeld University

2
Programme

• PreambleService-orientated libraries what does
  it really mean?
• Introducing ProSeBiCA Methodology, objectives
  and practical implementation, prospects
• The empirical surveys at Bielefeld University
• The idea generation step
• Design and results of the ACA survey (Adaptive
  Conjoint Analysis)
• Design and results of the CBC survey
  (Choice-Based Conjoint Analysis)
• Conclusions

3
Preamble

• Service orientation in academic libraries means
• Work must focus on the user.
• Service development must be in line with user
  needs.
• User acceptance will form a new (?) and vital
  criterion for valuation of libraries.

• Hence it requires
• to establish and improve customer relationship
• to create a profound knowledge about user
  (customer intelligence)
• to implement appropriate tools for library
  management

4
Introducing ProSeBiCA Methodology

• Conjoint Analysis marketing research method
• Aiming at
• ? the analysis of customer preferences on an
• empirical basis
• ? the simulation of potential decisions on the
  use of
• products / services
• A proactive method, looking at the future shape
  of products / services
• Widespread in commercial application

5
Introducing ProSeBiCA Methodology

• Technical terms
• Attribute and levela product / service and its
  potential forms
• Service offeringcombined levels of different
  attributes
• A priori segmentationrespondents subdivided
  according to pre-existing demographic criteria
• Benefit segmentationgrouping respondents
  according to the main benefits they look for in a
  product / service

6
Objectives of Project

• Adapt Conjoint-Analysis as a tool for the
  strategic development of academic libraries
• Develop an analysis and simulation framework for
  the whole range of library services
• First step framework for Bielefeld University
  as testing environment
• Second step general framework for academic
  libraries, guidelines for application of Conjoint
  Analysis
• Identify future relevant services, give insight
  what users may expect from their library

7
Practical Implementation

• Project funded by German Research Foundation
  (DFG)
• Carried out by Bielefeld University Library and
  the Chair of Marketing, Department of Economics
  and Business Administration at Bielefeld
  University
• Milestones

8
Prospects

• Co-operations with Sheridan Libraries of the
  Johns
• Hopkins University (JHU) and LibQUALTM
• Adapt ProSeBiCA into U.S. library perspective
• Integrate techniques offered by CAPM, LibQUALTM
  and ProSeBiCA
• Develop a homogenous tool and a portal as
  decision support system for library management

9
The idea generation step

• Sources of new ideas
• Secondary research(political papers, empirical
  studies from all over the world, Internet)
• Academic library user survey(Seminar on new
  library service development)
• Workshops with scientists and librarians(in
  Bielefeld, Cottbus and Kansas, USA)
• Statistical analysis of library usage
  data(detection of library usage types by means
  of NN)
• Academic library staff survey(OWL libraries)

• List including more than 250 new ideas
• for academic libraries

10
The idea generation step

• Example for new service ideas
• Video-mediated book-view
• 3D multi-mode search engine
• Webcam communication with librarians
• Information about the loan account via SMS
• Multilingual starter packages for foreign users
• Water dispensers in the library rooms
• Lockable carrels with extended communication
  facilities
• Personal university user card (multi-function
  card)
• Online publishing portal

11
The idea generation step
Classification scheme
12
Methodology of the empirical surveys

• Adaptive Conjoint
• Analysis
• Dynamical Question Generation
• Appropriate for a large amount of attributes and
  levels
• Four question types
• Attribute preference levels
• Attribute importance
• Paired-comparison trade-off questions
• Calibration concepts

• Choice-Based Conjoint
• Analysis
• Most used Online-Conjoint-Software
• Integrates the choice behavior of the consumers
• One question type Comparisons of complete
  library profiles

13
Design of the ACA survey
Example Attribute preference levels
14
Design of the ACA survey
Example Attribute importance
15
Design of the ACA survey
Example Paired-comparison trade-off questions
16
Design of the ACA survey
Example Calibration concepts
17
Design of the CBC survey
Example CBC choice concept
18
ACA survey at Bielefeld University

• Demographics and additional questions N 2122
• October 13th - December 10th 2004
• Provision of media (37 levels) N 579
• Learning and working environment (42 levels) N
  540
• Communication (24 levels) N 455
• Additional services (15 levels) N 548

19
ACA survey at Bielefeld University
Sample profile
20
ACA survey at Bielefeld University
Sample profile
Faculty membership
21
ACA survey at Bielefeld University
Most desirable services Conventional library
catalogue search Self-administration of the loan
account On-line post box Increase of the printed
media stock Optimized guidance system Upholstered
chairs More photocopiers Supported search
(performed jointly) Book information MyLibrary
personal profile Loan of yellow books
(presentation copies) Open-air area Water
dispensers Starter package University user
card Refereed on-line publishing portal
Least desirable services Conventional interior
design No on-line publishing No on-line search
support No search support by librarians No
personalization of library services Academic
search engine Recommendation system Radio-archive
Drive-in book lending Mobile navigation Children
s play area PDA (Personal Digital
Assistant) Library induction game Loudspeaker
system SMS notification On-line discussion
forum
22
ACA survey at Bielefeld University
23
ACA survey at Bielefeld University
Benefit segmentation Cluster analysis
(Ward) Provision of media Communication
Cluster 1 The involved and innovation-oriented
users (N 264) High usage likelihood Liking
different new services High importance of core
services
Cluster 1 The involved and innovation-oriented
users (N 184) High usage likelihood Strong
differentiation between different services Liking
new innovative services
Cluster 2 The traditional users (N
138) Rather low usage likelihood Low
differentiation between established
services Strong differentiation between new
services
Cluster 2 The less interested users (N
111) Low usage likelihood Low differentiation
between different services Liking some of the new
service ideas
Cluster 3 The online communication campaigner (N
103) High usage likelihood Strong
differentiation between different
services Preference of on-line communication if
possible
Cluster 3 The determined users (N 139) High
usage likelihood Strong differentiation between
different services High importance of core
services and some innovations
24
ACA survey at Bielefeld University
Benefit segmentation Cluster analysis
(Ward) Learning environment Additional
services
Cluster 1 The traditional users (N 214) Lower
usage likelihood than Cluster 2 (but not
uninvolved) Strong differentiation between some
of the services Preference for the current status
and, partly, for optimizations of the core
services
Cluster 1 The involved regular users (N
301) Rather high usage likelihood Comparatively
weak differentiation between services Open for
innovations
Cl. 2 The innovation-oriented users (N
131) High usage likelihood Distinct differences
between some of the services Explicit desire for
innovations and changes
Cluster 2 The innovation-oriented users (N
288) High usage likelihood Open for innovations
and changes Preference for optimizations of the
core services as well as for new
services Distinct differences between some of the
items
Cluster 3 The conservative users (N 29) High
usage likelihood Very strong differentiation
between some services Strict rejection of some
innovative alterations
25
CBC survey at Bielefeld University
Number of responses N 1672 May 04th - June
30th 2005
26
CBC survey at Bielefeld University
27
Conclusions

• Large samples covering all relevant user groups
• Lively interest / involvement of respondents
• Improved understandability of complex questions
  (due to textual and visual illustrations)
• Great importance of the conventional library as
  well as ofservices on the spot
• Willingness to deal with new service ideas
• No significant need of a plenty of extras
• Preference for services that ease the direct and
  effortlessaccess to information
• Partial confirmation the by results at the
  Brandenburg Technical University of Cottbus

28
Thank you for your attention!
Antonia Hermelbracht ahermelbracht_at_wiwi.uni-bielef
eld.de Tel. 0521 / 106-4883
Bettina Koeper bettina.koeper_at_uni-bielefeld.de Tel
. 0521 / 106-4057