Ethnographic claims require cautious transfer
Aliases: analytical generalization · transferability · thick description
What it is
Ethnography often supports analytical generalization: cases develop or test explanations about mechanisms, practice, and conditions rather than estimate a population proportion. Transferability depends on shared causal conditions, requiring thick description instead of the empty claim that findings cannot generalize.
Why it happens
Institutions, history, technology, and relationships jointly produce field observations. Some mechanisms recur elsewhere; others depend on local arrangements. Removing context makes superficial similarity look like causal equivalence. Necessary conditions, variation, and negative cases enable transfer judgments.
Studying it
Use maximum-variation or theoretical sampling to compare settings that challenge the account. Report case selection, thick description, negative cases, and inferential steps. Before transfer, compare contexts or run a small replication; do not attach prevalence language to purposive samples.
Where it stops holding
Caution does not reduce ethnography to unique stories: cases with explicit mechanisms and conditions can make strong theoretical contributions. Multiple sites are not population representative when they share the same access bias. Population prevalence or average effects require appropriate sampling and quantitative design.
Applying it
- Attach mechanisms, necessary conditions, and failure conditions to each claim.
- Compare roles, incentives, technologies, and institutions in the source and target settings.
- Pilot the critical mechanism before scaling an intervention.
- Label observed local evidence separately from proposed transfer.