Alliance effects on conflict more than double under a different control population
Evidence
In the set of politically relevant dyads there is almost no relationship (odds ratio=1.21) between alliance and dispute occurrence. In contrast, in the all-dyad population we ο¬nd an odds ratio over twice as high (2.60). β Gary Goertz, Social Science Concepts, 1521
Categorization
- XREF: Ties to the classic alliance-and-war debate; a concrete number showing how population choice drives it.