Data Analysis Tips and Tools: Exploring Qualitative and Mixed Methods approaches Fran Ackermann Lavinia Boyce Agenda Introduction Qualitative musings Mixed Method reflections Small group discussion Brief wrap up 1
Data Analysis Some thoughts from the qualitative camp Fran Ackermann Strathclyde University and Curtin Business School Why qualitative? interpretive techniques which seek to describe, decode, translate and otherwise come to terms with the meaning, not the frequency of certain more or less naturally occurring phenomena in the social world Van Maanen 1983 Get at richness and depth Awareness of different research method problems though! See McGrath s dilemmatic framework (quadrants recognising obtrusive/non-obtrusive research design & generalizable versus particular system behaviour) Many different methods (with associated data capture/analysis considerations) including (see Miles and Huberman) Ethnography Grounded Theory Case Studies Action Research Choose carefully which is appropriate to your question and your style Two Qualitative data capture/analysis techniques I am familiar with Cognitive or Causal Mapping Content analysis Fran Ackermann, BAM Doctoral Symposia, 2012 2
Where causal mapping has been used for research To elicit an understanding of what the interviewee means (through content & context) representation of perception To detect emergent properties within the structure To enable the interviewee to understand his/her world better To capture rich, deep, tacit data : experience, wisdom etc To improve interviewing capability If we know what statements mean and why they fit together as they do we can build a structure to pull a mess (complex problem) into a system of interacting issues 3
Analysis for research Reveal properties that can be the basis for comparing idiographic maps Manage complexity rather than reduce it e.g. other forms of mapping/repertory grids Detection of emergent patterns (individually and group wide) e.g. comparison of cognitive structures before and after a learning experience (Easterby Smith, 1980). identifying between monolithic, articulated or segmented clusters (Norris et al 1970). Exploring similarities/differences between value systems, central concerns, themes, options & loops Use of map analysis to manage complexity of maps 1 form Review structure (size, shape, ratios etc) Consider heads & tails (similarity/difference) Surface and test themes through: Identifying which concepts are the most busy Exploring the centrality of concepts Comparing themes Examining mutually exclusive sets (cluster analysis) Reviewing hierarchical sets explanation teardrops Carry out good auditing Use memo cards Use numbering/styles Use of Decision Explorer to undertake the analysis quickly and more accurately Consider using matrices to represent this material 4
Example of a map Fran Ackermann, BAM Doctoral Symposia, 2010 Where content analysis has been used To explore insights from a variety of qualitative text forms e.g. company reports, web questionnaires, interview scripts Think carefully about what is available To help with case studies for both explanatory and exploratory research (Eisenhardt) For both deductive and inductive research In combination with quantitative analysis examining in more depth a particular set of insights In combination with maps (triangulation) Fran Ackermann, BAM Doctoral Symposia, 2010 5
Analysis includes considering The a priori (typically from literature/rq) and emergent codes (from patterns) to help detect patterns analysis in its own right Careful and thorough procedures including Clear definitions of codes! Will require a cyclic approach Appropriate number of codes (and hierarchy) Avoiding revisions late on (particularly if using software) Care taken re inter-coder reliability test on colleagues! Avoid bias Identification of Inter-relationships between codes Counts frequencies of phrases (but not emotion or meaning) Assistance from software e.g. NVivo (electronic database able to manage coding, analysis and a range of different queries) 6