1 Qualitative Software Planning Seminar A comparative overview of various software packages which assist in the analysis of qualitative (textual or multimedia) data and/or facilitate the analysis of mixed methods projects. University of Birmingham 9 th November 2012 Dr Christina Silver CAQDAS Networking Project, University of Surrey, Guildford, UK Supported by the National Centre for Social Research (NCRM)
2 Background context Long established Training and Capacity Building programme Qualitative Innovations in CAQDAS (QUIC) research programme Methodological Innovations in Computational Support for qualitative research Integrating quantitative and qualitative data Analysis of (multi-stream) visual data Convergence of GIS technology with qualitative software
3 To encourage Our Mission Independent use of CAQDAS Critical evaluation of s/w tools Critical selection of s/w tools Critical selection of functions within s/w tools Constructive use of software to suit personal/ project/methodological requirements
4 Today. Basic overview of CAQDAS packages Review software developments Comparisons : methodological utility and practical usability Textual & multi-media analysis, mixed methods See also CAQDAS Networking Project Resources http://www.surrey.ac.uk/sociology/research/researchcentres/caqdas/support/index.htm
5 CAQDAS Definition Computer Assisted Qualitative Data AnalysiS Tools designed to facilitative qualitative approach to qualitative data including some of the following Content searching tools Linking tools Coding tools (not pre-requisite) Query tools Writing and annotation tools Mapping or networking tools Need not include all the above, may include others and may incorporate quantitative data and approaches
6 The Basic Idea of CAQDAS Packages Lewins & Silver (2007) Using Software in Qualitative Research : A stepby-step guide. Sage Publications. London
7 General similarities in software functionality Code and retrieve Text/Word searches Data organisation (e.g. socio-demographics) Searches of position of codes in data (e.g. Co-occurrence, proximity etc.) Writing tools (memos, comments etc.) Output of coded segments, results of searches etc..
Packages ATLAS.ti MAXQDA NVivo HyperRESEARCH QDA Miner Qualrus Transana Interact Observer Digital Replay System Dedoose TAMS Analyser
9 Potential role of software in the research process GENERALLY Overall - Improved access to data and thinking Project management tool Data storage, coding, searching, memo-ing Enhanced flexibility, readiness to ask questions/query, redo, rethink SPECIFICALLY Emphasis on particular tools/usefulness may depend on Research questions/aims Constraints Methodological & analytic approach Degree of familiarity with software
10 Choosing a CAQDAS package: Some questions to ask yourself Is there a package (and peer support) already available at your place of work? If not what sort of a budget have you got? What platform? MAC or PC? How much time do you have to learn the software? (Realistically, how many sophisticated tools do you expect to use?) What kinds and amounts of data do you have? What is your preferred style of working? Do you have a well defined methodology at the outset? Do you principally need to think thematically about the data? Coding less important? Principally need write notes in / at / about the data? Are you more concerned with the language, terminology used in the data? (e.g. the comparison and occurrence of words and phrases across cases or between different variables) Are you working individually or as part of a team?
11 Platform...Windows PC MAC users?...web based
Team working Transana Multi-user version
13 Team Working : real-time visual analysis in MiMeG
Team working Nvivo Server version
Team working Dedoose is a web-based tool
16 Code-based approaches Powerful code and retrieve functionality Many sophisticated tools reliant on early coding Within and between case analysis Inter-coder reliability Manipulate codes for alternative purposes Gathering tools (e.g. literature reviews, reflexivity) Pointers (e.g. colour coding) Generating output
17 Coding tools : accessing data Quick navigation, instant data availability Inductive, Deductive, Combined approaches Code collections, relationships Retrieval singly or together - enriches viewing of data Assigning attributes based on socio demographics etc. Similar devices in most packages Visual presentation varies
18 Memo tools reflecting upon data MEMO-ING work at text level Unpacking a statement Notes anchored to data Remind of insights Classify or collect memos into groups MAXqda, ATLAS.ti, QDAMiner (re memo flags in margin) Allows a mix of traditional and new tools Memo tools vary in appeal
Grouping tools factual features about data and respondents
Code retrieval : within and out of context
Data interrogation : query tool options
22 Key Differences Software architecture, ease of use, intuitiveness Textual data formatting (relevant to structured data) Handling and integration of literature Integration of geographical and time-based data Importation of social media type data Audiovisual data analysis Coding schema structures & margin display Closeness to data Software assistance in coding Support for non code-based qualitative approaches Support for text mining & quantitative content analysis Mapping tools Querying, searching & auto-coding possibilities Mixed methods support Collaborative working Data representation & output options
23 Key difference handling and integration of literature Relatively new functionality Importation of bibliographic database information Endnote RefWorks Zotero Evernote Enhances project management potential Enables fuller cross-referencing of data & analysis
Generating linked critical appraisals example in MAXQDA
Importing bibliographic info NVivo
26 Key difference integration of geographical and time-based data Relatively new area linking tools (MAXQDA, NVivo) coding and live navigation (ATLAS.ti) Combination of time and space (QDA Miner)
27 Geo-Linking in MAXqda
28 Geo-coding in ATLAS.ti
Geo-tagging in QDA Miner - the incorporation of time
Time-based analysis using the Track Viewer in DRS
31 Key difference support for the analysis of audiovisual data Use of CAQDAS less well established Status of visual data within a project Need for transcript key consideration Data representation key consideration Software bias towards code-based data analysis Some bespoke tools for video analysis
32 Synchronicity : multiple transcripts & video streams: example in Transana
33 Synchronicity : multiple transcripts & media types : Digital Replay System (DRS)
Outputting visual data Still images Moving images Recent improvements Viewability of annotations within software Options for outputting based on segmentation and coding Not video editing software Limits to output options Key to utility of CAQDAS for visual analysis where dynamic representations required Workarounds
35 Outputting the results of audiovisual analyses Link to html output Link to Word output
36 Key difference support for non Code-based approaches Analytic needs Maintaining holistic view of data Navigate without abstracting to conceptual level Track processes, interactions, contradictions etc Software tools Linking Annotating Networking Outputting
37 Navigating data without abstracting to the conceptual level : Example in ATLAS.ti
38 Working with hyperlinks : Example in ATLAS.ti
Key difference : Support for text mining & quantitative content analysis Many CAQDAS packages adding functionality Word frequency Tag cloud Cluster analysis QDA Miner fully integrated text mining with use of WordStat Clustering Dictionaries Multi-dimentional scaling Taxonomies Proximity plots etc.
Dendograms & Concept Maps in WordStat
Phrase Finder with KWIC in QDA Miner/WordStat
42 Key differences: mapping tools Each of these mapping tools individual qualities (All provide integration to data behind) Scribbled links NVivo and MAxqda = proactive links made between codes do not endure each map is a scribble - has no enduring effect Enduring links ATLAS.ti = proactive links to express relationship endure links made will recur in new networks Co-occurrence ATLAS.ti & MAXqda co-occurring codes, top-down interrogation
43 Mapping in ATLAS.ti
44 Mapping in NVivo
Mapping in HyperRESEARCH
Categorization tool - Qualrus
47 Key Differences : querying / searching Qualitative interactive cross-tabulations Differences in the range and flexibility of available searches between software packages Trade-offs between sophistication and usability Support for mixed-methods projects Alternative visual data representations The more complex the data organisation, the wider the possibilities for interrogation
48 Mixed Methods Integrating data Working with different media in parallel Incorporate materials directly Reference to externally held materials Developing synchronised written and audio/video transcriptions Integrating methods Importing quant measures about qual records Combining data types within an analysis Converting qual data into quant measures
49 Quantitative analysis of themes in the texts : example in NVivo
50 Code relations, Code matrix, Quote matrix in MAXqda
51 Automated coding by key-words Example in MAXqda
Hypothesis Testing Example : the 'Theory Builder' in HyperRESEARCH
Code Sequence Analysis Example in QDA Miner
54 Data Representation : charts Example in NVivo
55 Data Representation : charts Example in QDA Miner
Data Representation : charts Example in Dedoose
57 Methodological Utility Software as a project management tool Cyclical nature of qualitative data analysis Data Management & Research design Literature Review Meeting notes Data Representation and Presentation Code-based approaches Grounded Theory Thematic Analysis Framework Analysis/Template Analysis Non code-based approaches Discourse/Narrative/Conversation analysis Visual Anthropology/Ethnography/Photo elicitation Interaction Analysis Mixed Methods Integration of quantitative data and analyses
58 Flexible working Less linearity when using software Dip in and out of different functions Change your mind, go back and do something again, differently Build on something you ve already done CAUTIONS Being flexible doesn t mean inconsistent, patchy! Being flexible may take more time, not less! Using software doesn t guarantee more rigour in itself
60 References and key web sites Lewins A & Silver C (2007) Using Software in Qualitative Research: A Step by Step Guide, Sage Publications, UK di Gregorio, S & Davidson J (2008) Qualitative Research for Software Users, McGraw Hill, Open University Press, UK Silver C & Patashnick J (2011) Finding Fidelity : Advancing Audiovisual Analysis using Software, FQS 12(1), Thematic Issue: Is Qualitative Software Really Comparable? Silver C & Lewins A (2010) 'Computer Assisted Qualitative Data Analysis' in Penelope Peterson, Eva Baker, Barry McGaw (Editors), International Encyclopedia of Education, Vol 6, pp 326-334. Oxford: Elsevier Silver C & Fielding N (2008) Using Computer Packages in Qualitative Research, in Willig C & Stainton-Rogers W (eds.) The Sage Handbook of Qualitative Research in Psychology, London, Sage Publications. The CAQDAS Networking Project website http://www.surrey.ac.uk/sociology/research/researchcentres/caqdas/ The Online QDA website http://onlineqda.hud.ac.uk/index.php