Grounded Theory for ABM A Meeting of Methods

Grounded Theory for ABM: A Meeting of Methods? Professor Cathy Urquhart

My background in grounded theory Strong interest in qualitative data analysis, especially grounded theory, since 1995 Wrote a chapter on grounded theory in 2001 directed at postgraduate students which is widely used (Ed Eileen Trauth, Qualitative Research in Information Systems) Written in the Sage Handbook of Grounded Theory Written a commissioned book for Sage on grounded theory Senior Editor of MIS Quarterly, the top journal in information systems. MISQ has an impact factor of 4. 83, the highest of all business journals 2

Overview Background – a meeting of minds? What is grounded theory? What bridges can we build between GT and ABM?

Commonalities and Differences between GT and ABM Both are essentially inductive – they build a model or theory GT has historically used qualitative data ABM is quantitative, but is grounded in narrative observations 4

What is Grounded Theory Method? A qualitative research method. Co-originators sociologists Barney Glaser and Anselm Strauss (1967). An inductive approach to generating substantive theory that is ‘grounded’ in the data. Has clear directives on coding data. There are disputes about application and also the method. Many adaptations in many fields. © C. Urquhart 2000

Defining Grounded Theory 1 Dey 1999, Cresswell 1998 1. 2. 3. The aim of grounded theory is to generate or discover a theory. The researcher has to set aside theoretical ideas in order to let the substantive theory emerge. Theory focuses on how individuals interact with the phenomena under study. 6

Defining Grounded Theory 2 Dey 1999, Cresswell 1998 4. 5. 6. Theory asserts a plausible relationship between concepts and sets of concepts Theory is derived from data acquired from fieldwork interviews, observation, and documents. Data Analysis is systematic and begins as soon as data is available. 7

Defining Grounded Theory 3 Dey 1999, Cresswell 1998 7. 8. 9. Data analysis proceeds through identifying categories and connecting them. Further data collection (or sampling) is based on emerging concepts. These concepts are developed through constant comparison with additional data. 8

Defining Grounded Theory 4 Dey 1999, Cresswell 1998 10. 11. 12. Data collection can stop when no new conceptualisations emerge. Data analysis proceeds from open coding (identifying categories, properties and dimensions) through selective coding (clustering around categories), to theoretical coding. The resulting theory can be reported in a narrative framework or a set of propositions. 9

How Grounded Theory Approaches Coding 1. 2. 3. Open coding - the data is examined and coded at a word or sentence level. Selective coding – coding around emergent categories. Theoretical coding - coding for one or two ‘core’ categories and considering the relationships Note: Glaser version. Strauss and Corbin – open, axial, selective stages 10

Theoretical coding is important Many options for relating categories, one of which may be of particular interest to ABM colleagues. Glaser (2005) gives some wonderful nuances of causation in this theoretical code family. Causal family – a relative of the 6 Cs family, it includes several aspects: 1) bias random walk, 2) amplifying causal looping, 3) conjectural causation, 4) repetitive causal reproductions, 5) equifinality, 6) reciprocal causation, 7) triggers, 8) causal paths, 9) perpetual causal looping 11

Causal codes - 1 ‘Bias random walk’ – all variables are in a flux, ‘then on the introduction of a crucial variable … then of a sudden all of the variables fall into organisation’ ‘Amplifying causal looping’ – ‘consequences become causes, and one sees either worsening or improving progressions or escalating severity’ ‘Conjectural causation’ – it is not always easy to identify decisive causal combinations. ‘ 12

Causal codes - 2 Repetitive causal reproductions’ – a repeated action keeps producing the same consequences. ‘Equifinality’ – no matter what the causes and paths, the same consequence will occur. ‘Reciprocal causation’ – there is a similar interaction of effects or amplified causal looping. ‘Triggers’ – sudden causes that set off a consequence or set of consequences. ‘ 13

Causal codes - 3 Causal paths’ – used to intervene in changing or stopping a consequence. ‘Perpetual causal looping’ – a mathematical model, an ordered calculated growth of increased size based on a set temporal path. 14

Conclusion ABM and GT are closer than we might think? Both are in my view inductive methods Challenges are numerous Can GT help with isolating actions and agents? What would be the methodological sequence if we combined both methods? 15

Some other useful references Urquhart C, Lehmann H, Myers M (2010), Putting the Theory back into Grounded Theory: Guidelines for Grounded Theory Studies in Information Systems, Information Systems Journal, 20, 4, pp 357 -381 Urquhart C (2007) 'The Evolving Nature of Grounded Theory Method: The Case of the Information Systems Discipline' In: Kathy Charmaz and Tony Bryant (ed. ), The Handbook of Grounded Theory, Sage Publishers, 311 -331. Urquhart C and Fernandez W, Grounded Theory Method: The Researcher as Blank Slate and Other Myths, Proceedings of the Twenty Seventh International Conference on Information Systems, Milwaukee 2006 Charmaz K (2006), Constructing Grounded Theory, Sage Publications 16

Some Internet References www. groundedtheory. com http: //www. qual. auckland. ac. nz/grndrefs. aspx en. wikipedia. org/wiki/Grounded_theory_(Glaser) http: //gtm. vlsm. org/ 17

Open Coding - codes which become categories and properties of categories Selective Coding some codes become categories, some become properties, some relationships are made 1 -2 core categories ‘emerge’ Theoretical coding Finally, coding the data for the core categories. Building theory by considering relationships carefully. 18

Data excerpt Ok let me tell you about three different kinds of ways of using IT skills training […] There are organisations who provide generic IT skills [For example] learn to use Word, learn to browse the Internet, learn to use email […]. Those are the easiest ones to measure because you can have them [the people] a test […] you can issue a certification based on that Now, we have also found that it is the one that has the least retention and maybe the least usefulness for people to actually solve their every day problems. It maybe is the easiest one to implement but not the most productive one Another type of training that takes place is within an industry. Industry-specific training in IT skills. […] the training is tailored to certain people and training them in the specific skills that are needed in that industry 19

Open and selective coding – first try Possible Selective Codes Open Codes Generic Skills Training Agencies providing generic skills, word processing skills, web browsing skills, email skills, ease of measurement of generic skills, testing of generic skills, certification as outcome of generic skills, lack of retention of generic skills, lack of every day usefulness of generic skills, ease of implementation versus eventual value of skills Industry Specific IT Skills Training Job targeted IT training, Job gain as success measure 20

Open and selective codes – second try Possible Selective Codes Open Codes Types of skills training Industry specific IT training, Job targeted IT training, Agencies providing generic skills, word processing skills, web browsing skills, email skills, , lack of every day usefulness of generic skills, ease of implementation versus usefulness Benefits of skills training Job gain as success measure, Community empowerment, Benefits of local training, certification as outcome of generic skills 21

Analysis equals emerging dimensions of the research question One of the most interesting aspects of the interview is the idea that impact of IT skills training is hard to measure, and that the most impact is made by locally relevant training. So one emerging research question might be: How can we measure the impact of IT skills training? Another might be; How important is the context of IT skills training? 22

Theoretical Coding Types of skills training • Locally relevant training • Generic skills training • Industry based IT training Have different Benefits of skills training • Job gain as success measure • Community empowerment • Benefits of local training, • Certification as outcome of generic skills 23

Conclusion Grounded theory is a rigorous way of doing qualitative resarch Some researchers leverage the coding processes without producing theory and this is a legitimate use Good for rigour - you can be confident about what you produce Good for practice – enables us to build theories of what happens, instead of imposing theories from elsewhere Good for innovation – excellent where no theories exist -yet 24