Theory: role, generation and choice

1
Theory role, generation and choice

• Geoff Walsham
• Lecture 2 of Course on Interpretive Research in
  IS - Oslo University

2
Contents of Lecture 2

• Theory in research design/data collection
• Theory in data analysis
• Theory as final product
• Grounded theory
• Theory generation and choice
• Some Dos and Donts

3
Theory in Research Design and Data Collection

• Theory as a sensitizing device
• Example in Walsham (1995) use of Pettigrews
  content, context and process framework to look at
  IS strategy and implementation

4
Research Questions Generated

• What IS strategic approaches had the organization
  adopted? (content)
• What was the position of this organization in its
  sector and country? (context)
• How had the organization tried to implement its
  stated strategy and with what success? (process)

5
Example Structuration Theory(see Giddens 1984 -
also Walsham 1993)

• Three dimensions of meaning, norms and power
• Produced or reproduced in action
• Emphasis also on unintended consequences of
  intentional action

6
Interview Questions Generated(e.g. on ERP
systems)

• What do you understand this technology is aimed
  to do? (meaning)
• Does this technology require you to do your job
  in a different way? (norms)
• Who has required you to use this technology and
  what are their motives? (power)

7
Interview Questions (continued)

• What changes in your work, and that of others in
  the organization, has resulted from the use of
  this technology? (production).
• What has stayed the same? (reproduction)
• Did any unexpected things occur in the use of the
  technology? And what happened as a consequence?
  (unintended consequences)

8
Theory in Data Analysis

• Data collection followed by data analysis is too
  linear a model
• Theory can be valuable prior to data collection,
  during data collection and after data collection
• If, for example, you have used a theory to inform
  data collection, then it can also be used for
  data analysis

9
An Example of Data Analysis - Knowledge
Management Systems(Walsham 2005)

• Data on use of contact recording systems by
  salespeople in a pharmaceutical company
• Recorded details of their visits to doctors on
  database - available to others - sharing
  knowledge

10
Two Acts of Tacit Knowing(Polanyi 1969)

• Both the way we endow our own utterances with
  meaning and our attribution of meaning to the
  utterances of others are acts of tacit knowing.
  This represents sense-giving and sense-reading
  within the structure of tacit knowing.

11
Applied to the Pharmaceutical Example

• Understanding of interaction with doctors
  (cognitive sense-reading - person A)
• Putting the experience into words (performative
  sense-giving - person A)
• Interpreting the result (cognitive sense-reading
  - person B)

12
Difference Between TheFirst and Third of
These(Polanyi 1969)

• We may say that the observed meaning of an
  experience differs structurally from one conveyed
  in (words) the first meaning is immediately
  experienced while the second is only present in
  thought.

13
Theory as Final Product

• Theory can be the output of the work
• Popular with the top journals and examiners!
• But new does not have to be completely new -
  can be addition, reinterpretation etc

14
A Model of Basic Communication (Walsham 2005)
Action and reflection Person A
Re-presentation through voice, data, text,
diagrams etc. Person A
Reading of representation Person B
Action and reflection Person B
15
A Sociological Complement(Walsham 2005)

• What are the reasons for particular
  re-presentations (or deliberate
  misrepresentations) made by human actors?
• What is represented and for whom?
• Who requires particular representations, and for
  what purpose?
• What incentives (or disincentives) are provided
  for specific representations?

16
Grounded Theory

• Due originally to Glaser and Strauss (1967)
• Welcome emphasis on learning from the data rather
  than imposing a prior theoretical view
• I like the above emphasis but find the specific
  approach (and its later variants) too
  programmatic for my taste
• I prefer a balanced picture of learning from
  prior theory learning from the data

17
Some Donts ofTheory Generation and Choice

• Dont fix on one theory and then read nothing
  else
• Dont leave theory generation until the end of
  your study
• Dont dismiss a theorys value until you have
  read about it in depth

18
Some Dos ofTheory Generation and Choice

• Do listen to what others find valuable
• Do read widely on different theories
• Do choose theories which speak to you (not
  because they are the current fashion)