Doug Vogel
State Specially Recruited Expert
AIS Past President eHealth Research Institute Director
HIT School of Management
Case Study Research
1
Case Study
Useful
when “how” or “why” questions are being asked about events over which investigator has little or no control
Phenomenon often influenced by context
Exploration in depth (program, event, process, activities, one or more individuals, organizations)
Bounded by time and activity
Data collection sustained over time
2
Case Studies
Can
be single or multiple
Can be combined with other research methodologies e.g., surveys
Can precede or follow other methodologies in mixed method studies
Can embrace a range of levels of analysis e.g., individual, group, organizational, industries and even countries 3
Case Study Types
Exploratory
– Defining feasibility, questions and hypotheses for a subsequent study
Descriptive
(interpretive)
– Complete description of a phenomenon within its context
Explanatory
(positivist)
– Data bearing on cause-effect relationships
Use
Single or multiple case studies (for replication or contrasting)
4
Multiple Sources of Evidence
Documents
(reports, manuals, letters, meeting minutes, newspapers, websites)
Archival records (org. charts, budgets, computer files, transactions, emails, forums)
Interviews (guided conversations rather than structured queries – open-ended or focused)
Physical artifacts (buildings, office layouts, workspace, products, technology)
Direct observations
Participant observations
5
Design
Reflect
on goals and challenges
Create a conceptual framework
Identify research questions / propositions
Develop a case study protocol
– Consider embedded units
Identify
data sources
– Screen interview candidates
Think
through analysis
Develop a coding scheme
Do a pilot study
6
Case Study Protocol
More
than a questionnaire or instrument
Guide to carrying out the data collection including procedures and rules
– Overview of the project e.g., objectives
– Field procedures e.g., sources of information
– Case study questions e.g., data table “shells” and semi-structured questionnaires
– Guide for the report e.g., format for the data and presentation of documentation
Capture
the context as well as content
7
Embedded Units Design
Trimester 1
Trimester 2
Cohort Trimester Characteristics
Cohort Trimester Characteristics
(1A)
(1B)
(1C)
(2A)
(2B)
(2C)
Individual
Characteristics
Individual
Characteristics
Individual
Characteristics
Individual
Characteristics
Individual
Characteristics
Individual
Characteristics
8
Investigator Skills
Ask
good questions
Be a good listener
Be adaptive and flexible
Have a firm grasp of the issues being studied Be unbiased by preconceived notions
– Don’t just substantiate a preconceived position
– Be open to contrary findings
– Report preliminary findings to colleagues and encourage alternative explanations
9
Data Collection
Follows
a formal plan, BUT!!
Requires an inquiring mind during data collection, not just before or after
– Continually ask yourself why events or facts appear as they do
– May lead to immediate need to search for additional evidence
– One tentative answer can lead to a whole host of new questions
– Questions should eventually aggregate to lead to advanced understanding
10
Coding Process
Read text data
Divide text into segments of information
Code segments
Refine Codes
Collapse codes into themes 11
Deductive Coding
Start
with constructs associated with theory
Consider measures associate with constructs Create sub-codes more closely related to case characteristics
Identify text samples related to sub-codes
Combine coded text samples within measures and constructs
Closely examine un-coded samples for emergent insights/implications for theory
12
