Using Qualitative Data Software: An Introduction To NVivo

Using Qualitative DataSoftware: An introduction to NVivoGareth Harris

Overview Introduction to NVivoCritiquePossibilitiesPractical, the only way is doing

NVivo Three generations of CAQDAS: Text retrieval: lets users recover data where keywordsappear in the data Code-and-Retrieve: supports the division of text intosegments, attaching codes to the segments, andretrieving segments by code (or combination ofcodes). Theory-Building: emphasizes relationships betweenthe categories. Helps to develop higher-orderclassifications and categories, formulate propositionswhich fit the data and test their applicability, orvisualise connections between categories as an aid toconceptualization.

What can it do? Organisational tool: Manage data/ideasFacilitates accurate and transparent dataanalysis process and accountability.The ability to handle concepts as things, thus tomanage ideas, and explore their relations.Code-and-RetrieveTheory Building with the use of memo tools,queries, coding on live nodesVisualisationPotential for integration of quantitative/qualitative

What can’t it do? The software is a tool not a proxy for a validmethod of analysis.It cannot interpret the data for you? Similar toquantitative research.The question that needs to be asked is howmuch it impacts on the direction of datacollection & analysis?Neutrality of tools: analogy of a loomInappropriate for certain methods: Discourse,Narrative and Conversation Analysis

Criticism Quantification/mechanisation of qualitativedataFragmentation of data/ loss of narrative flowDistancing from dataDecontextualistionFetishisation of codingBias towards Grounded Theory

Creeping Quantification Can add transparency to the researchprocess but with the danger of empathisingstatistical criteria of validity and reliability thatmay be inappropriate to your research.Degeneration into searching for patternsOffers simple quantitative tools, word, textfrequency count. Matrices allow for aqualitative version of cross-tabulationsMechanisation of data analysis

Fetishisation of coding Dominance of code-and-retrieve to exclusionof other methodsIncreased reliance on coding and validationof patternsCoding becomes an end to itself instead ofgaining a higher level of interpretationAbstracting or adding complexity?Mapping the woods or cutting down trees?(Silverman: 2007)

Fragmentation/Decontextualistion Rapid coding allows us to identify repeatedpatterns, themes, concepts, categories etc.But does it allow us to understand howmeaning is constructed in conversationsbetween individuals or narrative structure.Early code-and-retrieve software wereheavily criticised for taking data out ofcontext. However this is also a problem forconventional manual methods.

Distance or closeness Tactile/digital divideQualitative analysis involves both closenessand distance.Too much closeness: ‘coding trap’ shutsdown abstraction and analysisAllows cognitive distance from the process ofanalysis: audit trail.

Grounded Theory NVivo seen to have an elective affinity withgrounded theory.The central goal is categorization –thediscovery, construction and development ofconcepts.Ease of coding/linking to memos & annotations/hierarchial node trees/coding-onFacilitates but does not guarantee theorydevelopment or coherence with particularmethod.

Code-and-Retrieve Allows rapid coding of large amounts of datacan use queries to detect patterns and rapidlycodeCoding system in which researcher’s evolvingknowledge about the data is stored in nodeswhich retain the link to the source dataIterative process of recoding through codingon live nodes: ‘Coding-on’.

Queries Allows code-based theory-buildingExpands ways of asking questions aboutcoding and sourcesSaves and reuses the product of queriesResults can be coded: System closureAllows searches that drives rather than endsenquiryCoding that provides data for more analysis.

Cantle report

Shared Futures

Importation of demographic data Classification tables allow for importation ofdemographic dataMatrix queries allow the researcher tosystematically look at differences amongstparticipants, how participants constructmeaning differently according to variousattributes, demographics but also sequences.Can compare coded text by attributes ofcases

Visualation Increases accessibility of ideasExcellent exploratory tool to rapidly identifykey conceptsCan be combined with coding to rapidly codelarge amounts of dataUnderstanding of relationships and modelbuilding

Tree maps

Charts but quantification?

Beyond coding 7 types of analysis which NVivo can facilitate: Constant ComparisonClassical ContentKeyword in contextWord countDomain analysis: Symbols, the cover term, theincluded termTaxonomic analysisComponential analysis (Leech,N: 2001)

Overview Nvivo WorkspaceImporting sourcesDownloading word documents fromwintersresearch.wordpress.comImporting .doc into NVivoOpen coding reading of transcriptsQueries

Resources/Methodological aqdasprimer.php Very CriticalSpecial Issue: International Journal of Social ResearchMethodology, 2002, Vol. 5, (3)Bringer, J: Using Computer-Assisted Qualitative DataAnalysis Software (CAQDAS) to Develop a GroundedTheory Project. Qualitative Research, 2004, Vol 4, (2)Online QDA at http://onlineqda.hud.ac.uk/

Practical See Nvivo 9 Getting Started free 30 daytrial! athttp://www.qsrinternational.com/#tab youBazely, P (2007) Qualitative Data Analysiswith Nvivo (Sage)CAQDAS Network rchcentres/caqdas/

NVivo Three generations of CAQDAS: Text retrieval: lets users recover data where keywords appear in the data Code-and-Retrieve: supports the division of text into segments, attaching codes to the segments, and retrieving segments by code (or combination of codes). Theory-B