Within-case and cross-case analysis

Description

First builds an understanding of each case as a whole, then compares cases systematically to find patterns, configurations and explanations that hold across them.

Variable-oriented synthesis pools quotes across participants and loses the configuration of conditions within each case. Within-case analysis first constructs a coherent account of each case (a person, a clinic, an account), often as a case summary and display. Cross-case analysis then compares cases using strategies such as stacking comparable displays, partitioning cases by outcome, building meta-matrices, and identifying configurations of conditions associated with outcomes. Yin's replication logic treats each case like an experiment: literal replication predicts similar results, theoretical replication predicts contrasting results for predictable reasons. Rigour comes from standardising within-case write-ups so they can be compared, from case selection that supports replication logic, and from explaining every case, including deviant ones.

Attributes

Loss type
Siloed
Stage
Definition
Effort
Weeks

Steps

  1. Define the case unit (a user, household, clinic or organisation) and select cases to allow literal and theoretical replication, typically 6 to 20 cases.
  2. For each case, write a standardised within-case summary of 2 to 4 pages and build a within-case display, such as a time-ordered matrix or causal network.
  3. Choose cross-case dimensions from the research questions and build a meta-matrix with cases as rows and dimensions as columns.
  4. Partition and sort cases by outcome (for example adopters versus non-adopters) and look for conditions that consistently differ between groups.
  5. Identify configurations of conditions that co-occur with outcomes, and check each deviant case to explain or bound the pattern.
  6. State cross-case propositions with the number of supporting and contradicting cases, and link back to within-case evidence.

Sources

  • Qualitative Data Analysis (Miles & Huberman) via Sage Handbook
  • Cambridge Handbook of Research Methods
  • The Inquiry Series, Ch 29 — Miles & Huberman — Systematic Qualitative Data Analysis
  • Miles, M. & Huberman, A. M., Qualitative Data Analysis: An Expanded Sourcebook, 2nd ed. (1994)

Also known as

  • Case-oriented analysis
  • Cross-case synthesis
  • Multi-case analysis

Categories

  • User Research
  • Qualitative Analysis & Synthesis