Docstoc

Chapter 13 Analyzing Qualitative Data

Document Sample
Chapter 13 Analyzing Qualitative Data Powered By Docstoc
					Ashley Webb
 Paul Sedor
Will Vonier
 Sonia Ortiz
 Qualitative data defined
 Examples of Qualitative data
 To be useful, data analysis must be conducted
 Two types of qualitative data analysis include
  deductive and inductive
 Computer aided qualitative data analysis software
  (CAQDAS)
 CAQDAS includes NVivo, ATLAS.ti, N6, and
  HyperRESEARCH.
 Software for quantitative data is universal, use of
  CAQDAS for qualitative data is not so widely practiced
 While ‘number depends on meaning,’ (Dey) it is not
  always the case that meaning is dependent on number.
 ‘The more ambiguous and elastic our concepts, the less
  possible it is to quantify our data in a meaningful way.’
  (Dey)
 Qualitative data is associated with concepts and is
  characterized by its richness and fullness based on
  your opportunity to explore a subject in as real a
  manner as is possible.
Quantitative Data                        Qualitative Data
•Based on meanings derived from          •Based on meanings expressed through
numbers                                  words
•Collections results in numerical and    •Collection results in non-standardized
standardized data                        data requiring classification into
                                         categories
•Analysis conducted through the use of   •Analysis conducted through the use of
diagrams and statistics                  conceptualization
 Data grouped into categories before it can be
  meaningfully analyzed, otherwise the most that may
  result may be an impressionistic view of what they
  mean.
 Analyzing qualitative data not an ‘easy option.’
 Data analysis should be consider at time of
  formulating proposal.
 Process of analyzing qualitative data begins at same
  time as you collect the data and continues afterwards.
 Chapter 8 – secondary data
 Chapter 9 – observations can be video-recorded
 Chapter 10 – transcribe recordings and notes to
  prevent loss of data
 Chapter 11 – open questions used to collect qualitative
  data
 Chapter 12 – analyzing quantitative data
 Now in Chapter 13, we focus on the conversion of
  qualitative data to word-processed text since this is the
  form we will most likely use to analyze our qualitative
  data.
 Non-standardized interviews
   interview is normally audio-recorded and transcribed,
    that is, reproduced as a written account
 Interviewer interested in not only what participants
  said, but the way it was said
 Alternative ways of reducing the time needed to
  transcribe audio-recordings along with the potential
  problems that could be involved. (pg. 475)
 Data cleaning
 Ensure factual accuracy
 Have you thought about how you intend to analyze
  your data and made sure that your transcription will
  facilitate this?
 Have you chosen clear interviewer and respondent
  identifiers and used them consistently?
 Have you saved your transcribed data using a separate
  file for each interview?
 Have you checked your transcript for accuracy and
  where necessary, ‘cleaned up’ the data?
 Forms of textual data
   email interviews
   electronic versions of documents, including
    organizational emails and web-based reports.
 Spend time preparing it for analysis. Involves making
 sure data is:
      Suitably anonymized by using separate codes for yourself
       and different participants
      Appropriately stored for analysis, for example one file for
       each interview, each meeting’s minutes or each
       organizational policy
      Free of typographical errors that you may have introduced
       and where these occurred have been ‘cleaned up’
 No standardized approach to analysis of qualitative
  data
 Several different strategies to deal with the data
  collected These strategies are divided into four main
  categories:
       Understanding the characteristics of language
       Discovering regularities
       Comprehending the meaning of text or action
       Reflection
 First two categories - analytic strategies (deductive)
 Second two categories – no predetermined structures
  (inductively)
 The general set of procedures listed below elaborate on
 the aspects of qualitative analysis and involves the
 following activities:
   Categorization
   ‘Unitizing’ data
   Recognizing relationships and developing the categories
    you are using to facilitate this
   Developing and testing theories to reach conclusions
 Classifying data into meaningful categories
• Sources to derive names for these categories:
o  Utilize terms that emerge from the data
o They are based on the actual terms used by the

  participants
o They come from terms used in existing theory and

  the literature
• Categories have an internal and external aspect
Attach relevant ‘bits’ or ‘chunks’ of the data,
which refers to units of data, to the appropriate
category or categories. This can be done using
CAQDAS, the manual approach, or by indexing
categories.
Example:
 Search for key themes and patterns or relationships in
  the rearranged data
 Revise categories and continue to rearrange the data
 May subdivide or integrate categories as ways of
  refining or focusing the analysis
 Keep an up-to-date definition of the categories
 Development of hypotheses to reveal patterns within
  the data and recognize relationships between
  categories
 All the relationships need to be tested
 Test the hypotheses or propositions that inductively
  emerge from the data
 Development of valid and well-grounded conclusions
 Allows to recognize important themes, patterns and
  relationships as you collect data
 Allows to adjust the future data collection
 This process has implications for the way in which you
  will need to manage your time and organize your data
  and related documentation
 It will be necessary to arrange interviews or
  observations with enough space between them
 Summaries, including those for interviews,
  observations, and documents, and also,
  interim ones
 Self-memos
 A researcher’s diary
 Summary of the key points that emerge from
  the written part of the notes
 Allows to identify relationships between
  themes
 Useful to make comments about the
  person(s) interviewed or observed, the
  setting in which this occurred and anything
  occurred during the interview
 Allow to record ideas that occur about any
  aspect of the research
 May vary in length from a few words to one
  or more pages
 Can be written as simple notes
 Should be filed together, not with notes or
  transcripts
 May be updated as research progresses
 Record ideas and the reflections of these
 Act as an aide-memorie to the intentions
  about the directions of the research
 Allows the identification of the development
  of ideas and the way in which the research
  ideas are developed
 Provides an approach that suits the way in
  which you like to think
 As mentioned earlier, data collection is either:

      Deductive -use existing theory to guide your approach


      Inductive - develop a unique theory that is base solely upon
       your data.
Using a Theoretical or Descriptive Framework:
  If you use existing theory to create your research question –
  you may also use existing info to build your frame work and
  organize you analysis of data.
Disadvantages:
 closing the research prior to finding the best conclusion
 influenced by the social activities of the participants.
Advantages :
 Link research to an existing body of knowledge
 Help get started
 Provide an initial analytical framework
Exploring without a predetermined theoretical or
descriptive framework:
Collect data then analyze them to see if there are any
patterns or themes that develop.
 May be difficult
 May not lead to success for inexperienced researchers
 There is no clearly defined framework in the beginning
 You will need to analyze the data as it is collected
 Then develop a conceptual framework as a basis for the rest of
  your work – also called Grounded approach
 This method has proven to be resource intensive as well as
  requiring a large amount of time.
Most research combines elements of both the inductive
and deductive approach
Pattern matching:
  Predicting outcomes based on theories to show what you
  are expecting to find as a result of your research.

This approach involves:
 Developing a conceptual or analytical framework that is
  based on existing theory
 Test the framework to explain your data that is collected
 May not lead to success for inexperienced researchers

  If your predicted outcomes match what is shown through
  the conceptual framework then this leads to an explanation
  thus discounting any questions of validity.
Explanation Building:
  Build an explanation while collecting data and analyzing
  rather than testing a predicted explanation.
  Similar to grounded theory but is designed to test a theory
  rather than generate a grounded one. And uses the
  following procedural stages:
   Devise a theoretical based hypothesis
   Collect data and compare the findings to the theoretical
    based hypothesis
   If necessary amend the hypothesis according to any new
    findings
   Collect more data and compare the findings to the revised
    hypothesis
   Continue the process until you achieve a sufficient
    explanation
Impact of a deductive approach on the analysis process:
 You will still follow the general process of analyzing
  qualitative data.
 Your hypothesis will still need to be rigorously tested,
  but using predicted explanations should force the
  answer to your research question to be more specific.
 This will depend upon how thorough you used the
  existing theory and framework as well and the
  appropriateness of the hypothesis and conceptual
  framework that emerges from your data.
Several inductively based analytical procedures to analyze qualitative data
  such as:
      Data display and analysis
      Template analysis
      Analytic induction
      Grounded theory
      Discourse analysis
      Narrative analysis
Many of these combine elements of inductive and deductive approaches.
Reasons to use inductive analysis:
    Looking to generate a direction for future work
    Research scope is constrained by restrictive theoretical propositions
     that are not governed by the participants personal views
    The theory may point to later actions because it was developed as a
     result of the setting in which the research was conducted
    The theory may be of a nature general enough to be applied in other
     contexts
The inductive approach should not be used as a means to avoid in depth
  preparation prior to starting your research project.
Data display and analysis:
  Based on the book by Miles and Huberman that
  focused on the process of ‘doing analysis.’
Process contains 3 sub-processes:
   Data reduction - clarifying and simplifying the data
   Data display – organizing and grouping data into visual
    displays so that patterns and relationships can be
    recognized
   Drawing and verifying conclusions
 The exact procedures to follow in these sub-processes
 are not specified and can be amended according to
 what is appropriate within the context of your project.
 Template analysis:
   A procedure to analyze qualitative data based on the work of
     King. A template is a list of themes that are revealed by the
     data that has been collected.
 A template is a list of codes/categories that emerge from
  the collected data.
 Template analysis is less structured and prescriptive than
  the grounded approach and allows more flexibility when
  altering to fit the needs of your specific research project.
 When altering the codes or adding new codes you should
  be careful and be aware of its impact on previous coding
  activity. Also you must document your reasons well.
 Template analysis:
    Ways in which a template may be revised:
    1.      Insertion of a new code into the hierarchy as the result of a
            relevant issue being identified through data collection for which
            there is no existing code
    2.      Deletion of a code from the hierarchy if it is no longer relevant
    3.      Changing the scope of a code (altering its level within the
            hierarchy)
    4.      Reclassifying a code to a different category
        The template may continue to be revised until all of the data
         has been collected, analyzed and coded to satisfaction.
        The template approach can help select points that need to be
         addressed further during the course of your research as well
         uncover themes and issues that may not have been apparent
         from the onset of the project.
 Definition
 Less defined explanation of phenomenon
 Case Study process
 Methods of case studies
   In-depth Interviews
   Observations
   Combination
 Criticisms
 “Theory Building”
 Stages
   Open Coding
   Axial Coding
   Selective Coding
 Implications
 Analysis of language
 Different discourses / norms
 Three-dimensional Analytical Framework
   Text
   Discursive Practice
   Social Practice
 Disadvantages
 In-depth Interviews
 Like a story
   Beginning
   Middle
   End
 Structural Elements
 Two usages
 Count frequencies of events, reasons, or references
 Ignores nature and value of data




 • Functions of software
 • Exploring latest versions

				
DOCUMENT INFO
Shared By:
Categories:
Stats:
views:54
posted:9/20/2011
language:English
pages:36
pptfiles pptfiles
About