Qualitative research

1
Qualitative research

• Sheila Webber
• University of Sheffield, Department of
  Information Studies
• s.webber_at_sheffield.ac.uk
• August 2002

2
Outline

• Why qualitative research?
• Access
• Collection methods interview, focus group
  observation
• Approaches case study ethnography Action
  research
• Analysis
• Service quality in libraries
• Summary

3
Why qualitative?

• Reveal hidden, unsuspected issues (may be asking
  wrong questions on a questionnaire!)
• Exploring attitudes, emotions, sensitive issues,
  opinions, conceptions
• Exploring context, relationships, processes
• What is objectivity anyway?
• Quantitative qualitative methods can be
  effective
• Triangulation
• Using data from one exercise to inform another

4
Access and understanding

• Need access to your population
• If do not have access, difficult to get good
  sample, or enough time with respondent (or any
  response at all)
• Typical access problems surveying non-users of
  services surveying businesses
• Particularly e.g. for case study, important to
  have pre-understanding of culture, language etc.

5
Collection methods

• Interviews
• Focus groups
• Observation

6
Collection methods interview

• Good for open ended questions, probing and
  exploring in depth, gaining understanding
• Good where want impression of body language,
  intonation, whole person, context and setting
• Improved understanding of questions by
  interviewee
• Expensive (time/money collecting and analysing)

7
Interviews

• Need good interpersonal skills
• create rapport
• active listener
• self-controlled and neutral
• adaptive
• unobtrusive efficiency control of interview
• Good planning, structure
• Closed or open questions
• Probes
• Structured, unstructured or semi-structured

8
Interviews

• Channels face-to-face telephone email video
• Computer Assisted Personal Interviews Computer
  Assisted Self Interviews
• Email interview could build up panel
• Channel choices body language important? Need
  to survey dispersed population? Will channel
  encourage frank speaking? Each has pros and cons

9
Focus groups

• Good for how which and why questions
• As group can interact, more options for providing
  insights
• Small groups (e.g. 6-8)
• Moderator and observer
• Taping is recommended
• Setting should be carefully planned and
  controlled appropriate to aims, topic and
  participants

10
Focus groups

• Activities which can engage all the group to
  avoid the lets go round and each say what they
  think of this syndrome
• Few questions, but well planned activity to
  maximise interaction and encourage participation
  from all on each question

11
Focus groups

• Use of internet chat or bulletin boards they
  have some uses (good when participants dispersed,
  housebound etc.) but also some drawbacks (e.g.
  more difficult to have quick interaction between
  people)
• Focus groups can help generate questions for
  surveys, or can be used to probe issues arising
  from analysis of quantitative research

12
Observation

• May be silent or involve neutral questioning
• May involve active participation of researcher in
  events
• May involve video, audio taping, examining web
  logs etc.
• Important for website usability studies (e.g.
  using thinking aloud protocols) also e.g. how
  do people use reference collections
• Often useful to combine examination of data (e.g.
  transaction logs) and observation of what people
  do

13
Approaches

• Case study
• Ethnography
• Action research
• Phenomenography

14
Case study

• An enquiry (how, why questions) investigating
  a process, phenomenon
• Has clear boundaries
• Examining events, people, processes in a specific
  context aim for holistic understanding
• Cases need to be selected carefully but do not
  have to be typical
• May use combination of methods (desk research,
  observation, interviews etc.)
• May be linking to thoery trying to link
  cause/effect

15
Case study

• Exploratory studies may generate material for
  other types of research
• Could be useful if your customers are
  organisations (e.g. how does company X use
  information in making decisions about Y)
• Could be useful
• to present to funding bodies to justify service
  changes (gives tangibility and personality)
• to guide service changes
• to explain benefits in publicity

16
Ethnography

• Settings carefully observed and described (unlike
  specific how/why of case studies)
• Population observed and encouraged to reveal
  "realities" of their lives (although question as
  to whether people can reveal "reality")
• Researcher aims to integrate into setting without
  disrupting or affecting what is going on

17
Ethnography advantages

• May reveal issues that the researcher may not
  have considered (something you would never have
  asked a question about)
• Can contribute information on attitudes, on
  "how", "why", "when"
• Vivid pictures of potential library and
  information users may appeal to decision makers
  help them to understand issues better

18
Action research

• When you want to change something, be part of the
  process, but observe
• May be testing hypothesis
• Process
• Identify something specific you think needs
  investigating / changing
• Plan actions you can be part of the action
• Record and reflect as you go along
• Get feedback from colleagues, customer, managers
  etc.
• Review, report if appropriate
• Plan changes
• Start cycle again

19
Action research advantages

• Can be done alongside things you wanted to do
  anyway but you need to plan ahead
• Plan time for reflection
• Involving other people (customers, colleagues) in
  reflection/critique about what is going on
• Go through the cycle again so you can apply what
  you've learnt from the first iteration
• Very applicable to user education

20
Phenomenography

• Investigating peoples conceptions of a
  phenomenon (e.g. conceptions of information
  literacy)
• About 20 indepth interviews with only a few
  questions and lots of probing
• Aim to identify variation in conceptions
  textual descriptions of categories plus a
  diagrammatic representation of how different
  conceptions relate to one another

21
Phenomenography

• Cannot assume results can be generalised to
  another population
• Tells you where people are you are attempting
  to understand their viewpoint thoroughly
• Useful if you are interested in really
  understanding your population and addressing
  diversity of interest/ need
• Interviews and analysis are very time consuming

22
Analysis

• Often involves transcription of audio- or
  videotapes - dull and/or expensive!
• Analysing text (or other evidence) for themes,
  concepts, variation
• Manual analysis still very important
• Ethnographic approach involves narrative,
  reflection of researcher may be part of
  narrative, rich in quotations

23
Analysis

• May be more emphasis on coding concepts,
  identifying relationships hierarchies between
  concepts, textual analysis
• May go on to identify categories e.g. with
  pen-pictures of different kinds of customer (e.g.
  Early Adopters Fast Forwarders Mouse Potatoes
  Techno-Strivers New Age nurturers Sidelined
  Citizens. (Forrester Research 1998)
• Latter useful for market segmentation

24
Analysis

• Qualitative analysis software can aid process
  (though does not replace intellectual effort)
• NUDIST and Nvivo
• ATLAS/TI

25
http//www.atlasti.de/Atlnet.html
26
http//www.atlasti.de/Atlnet.html
27
http//www.qsr.com.au/products/nvivo_details.html
28
Service quality in libraries

• Not particularly qualitative research..
• Uses SERVQUAL approach looking at gap between
  expectation and performance
• Large scale exercises in USA (LIBQUAL) and
  Australia in university libraries

29
Summary

• Time consuming but worthwhile
• Helps you understand your market, have new
  insights and avoid assumptions
• Can help you segment your market, target your
  products more effectively, focus better on
  benefits, talk their language also may produce
  striking pictures which bring your services to
  life for funders