Stupid Case Study Shit

Case Studies:
→real life context
→more points of interest beyond data points
→multiple data sources ⇒ triangulation of data (cross-verification) = better reliability
	→observation
	→interviews
	→outright data

Case Study Phases
→Design
→Prep Data Collection
→Data Collection
→Analysis of Evidence/Data
→Report

☆Design
→Prep questions
→Theoretical propositions (theories about software design)
→Unit of analysis

	→Single case
	→Multiple cases

	→Holistic = Single unit of analysis
	→Embedded = Multiple units of analysis

→Logic between data and propositions
→Criteria for interpretation
	→Validity
		→Internal Validity
			→Causality within the case study; proper procedure for the case study
				→Pattern-matching
				→Explanations
				→等々
		→External Validity
			→Consistency of the theory between different cases
				→In case of single case: use theories sourced from other reliable sources (literature, whatever)
				→In case of multiple cases: replication logic (sharing information)
	→Reliability
		→Operations (procedures, etc.) can be repeated with same results
		→Use a case study protocol
		→Case study database

☆Prep Data Collection
→Creating a case study protocol

	→Overview
		→Objectives
		→Issues
		→Topics
	→Field Procedures
		→Credentials
		→Access privileges
		→Information sources
			→People
			→Documents
			→等々
	→Case study questions (things to keep in mind)
	→Guide to report
		→Outline
		→Format for narrative

Things to note:
→Ask good questions, interpret answers
→Listen
→Adaptability and flexibility
→Firm grasp of issues
→Unbiased

☆Data Collection
Sources:
→Documents
	→Letters
	→Memos
	→Memoranda
	→Yeah.
→Archives
	→Records
	→Lists of stuff
	→Survey data
	→Yeah.
→Interviews
	Types:
	→Open-ended
	→Focused
	→Structured
	→Survey
→Observation
→Observation with participation
→Physical artifacts

Always:
→Use multiple sources
→Use a case study database/repository
	Made of:
	→Notes
	→Documents
	→Tabular stuff
	→Narratives
→Maintain a chain of evidence

Types of triangulation:
→Data
→Investigator
→Theory
→Methodological

☆Analysis of Evidence/Data

General	Strategies:
→Defend theoretical propositions
→Compare with rival propositions
→Develop case description

Specific Strategies:
→Pattern Matching
→Explanation Building
→Time Series Analysis
→Logic Models
→Cross-case Synthesis

Principles of Analysis:
→Attend to all evidence
→Address any and all rival interpretations
→Address most significant aspect(s) of case study
→Use prior, expert knowledge

☆Report
What people want to know:
	Academe
		→Connections
		→Findings
		→Previous theory/research
	Nonspecialists
		→Descrption of the situation
		→Implications of actions in the situation
	Thesis Committee
		→Mastery of theory and methodology
		→How the study was conducted
	Reseach Funders
		→Significance of case study findings with regards to academic of practical terms
		→Rigor of how the study was conducted

Possible Formats:
→Classic single-case
→Multiple-case version
	→Multiple narratives
→Series of research questions and answers
→Cross-case analysis

Composing the Report:
→Start during analysis
→Determine level of disclosure of data and sources
→Controversial topic or not?
→Consequences on other similar studies
→Determine if case is "ideal"

Things to do:
→Make sure it's reviewed
	→Peers
	→Participants
	→Informants
	→Other significant parties

Good case studies are:
→Significant
	→Unusual
	→Interesting
	→Important to issues
→Complete
	→Full of info
	→Analytical and stuff
	→Covers all angles/perspectives
→Alternative perspectives
→Sufficient Evidence
→Engaging to the reader