Qualitative Data Analysis : An Introduction

1
Qualitative Data Analysis An Introduction

• Carol Grbich
• Chapter 21
• An overview of qualitative computer management
  programs

2
Qualitative Compter management programs

• Qualitative computing packages comprise three
  main types
• code and retrieve
• theory generation
• content analysis
• These programs provide useful tools but you need
  to be aware of their pitfalls and limitations.

3
1. Systematisation Code and Retrieve programs

• Ethnograph 6 large file capacity
• Ethno 2 - hierarchical links for events
• SuperHyperQual 1.4 automatic and
  semi-automatic coding

4
2. Theory Generation programs

• ATLAS.ti 6.2
• Rocket FolioVIEWS
• HyperRESEARCH 2.6
• MAXQDA 10
• NVivo 9
• NUDIST N6

5
3. Content analysis

• These packages are useful tools for breaking into
  text. Most have the capacity to undertake
• word frequencies indicating how often each word
  occurs in a document
• category frequencies where synonyms are grouped
  into categories and the program shows how many
  times each category occurs in the document
• KWIC (key word in context) which displays, in
  alphabetical order, each word together with a
  number of words on either side to provide
  information on its context in the document
• cluster analysis where groups of words can be
  identified as being utilised in similar contexts
• co-occurrence of pairs of words. The more
  sophisticated of these programs are developing
  the capacity to attempt semiotic analysis and
  often include cultural grammars.

6
Content analysis programs

• asksam 6
• QDA Miner 3.2
• Textpak 7
• Wordcruncher 7.1
• Wordstat 6.1
• ZyIndex
• Visual data
• C Video
• Transana 2.4

7
Qualitative computing packages continuing
concerns

• The tools of computer programs
• Tools constructed for a particular program must
  inevitably impact upon the data..
• 'Reality' has to be segmented, truncated and
  textured to prepare data to 'fit' a particular
  form of programming..
• These processes also promote procedural thinking
  and high frequency logic, and have the capacity
  to enhance the distortion of time.
• The framing of knowledge
• The sequential and procedural approaches which
  are intrinsic to individual programs must
  inevitably structure thought processes and
  texture the data in particular ways.
• The act of enclosure within framing processes
  confines the data and separates out the data from
  both the researcher and the context.

8
Qualitative computing packages continuing
concerns

• The texturing of reality
• Views of reality are formal, discrete,
  reductionist, algorithmic, sequential,
  deterministic, mechanical, computational,
  atomised, digitalised, logical and rational
• The representations of 'reality' that are
  produced are based primarily on discreet objects,
  not thick interpretation
• Impact on knowledge
• The gendering of technology through 'masculine'
  logic has the potential to distance women from
  involvement.
• There is considerable potential for
  decontextualised fragments to be hammered into a
  shape convenient to the purposes of the
  manipulator
• The capacity of programs to handle larger and
  larger data sets frees up researcher time but
  encourages more and more data to be collected.,
• Garbage In, Gospel Out'

9
Qualitative computing packages continuing
concerns

• Communication - networking
• these structured forms of communication (person
  to computer, computer to computer) bear little
  resemblance to more usual processes of
  collaborative research.
• Reification (or the glorification of the status
  of computer management)
• Where computer management is seen as being the
  only option
• Once codes have been developed, they have a
  tendency to become objectified or regarded as
  variables to be looked at in terms of frequency
  of occurrence.
• Quantitative interfaces
• The quantitative interface of SPSS imposes
  additional frames on the data

10
Users' comments

• Idiosyncracies
• Many of the programs are idiosyncratic in nature,
  originally developed by individual academic
  researchers to help manage their own databases
• Time
• Considerable time is required for the processes
  of setting up the data and learning to operate
  the program efficiently
• Decontextualiasation of data
• Computing packages targeted at qualitative
  research tend to emphasise the common and shared
  properties of a large number, rather than
  variations in the minutiae of detail of small
  numbers
• Etic view
• the etic (outsider, 'objective') rather than the
  emic (insider, 'subjective') approach to
  interpreting meaning is fostered

11
Users' comments

• Best practice
• multiple analytic approaches (manual plus
  computer managed) enhance dependability
• it is crucial to be a competent methodologist
  prior so you have control over the process. These
  programs are tools for the experienced researcher
  to use wisely, not substitute tools to cover lack
  of expertise
• Audiovisual data
• Are viewed as blunt tools in a fairly
  primitive state
• Which program is better?
• MAXQDA with its integrated code, comment and memo
  functions supports the interrelationship among
  data, coding and memoing far better in regard to
  design, ease of use and theory generation,