Evaluating the imitation game as a method for comparative research: a replication study using imitation games about religion

The Imitation Game is a new method and, as such, it is important to show that its results are plausible and replicable. We tested this by conducting Imitation Games on religion in a range of European countries, returning approximately 12 months later to repeat the research. The idea was that non-Christian members of strongly Christian countries would find it easy to pass as members of the practicing majority because Christian beliefs and practices would be ubiquitous. In more secular countries, the expectation was that non-Christians would find it harder to pass as Christian because religious practices are less visible. We show that, despite some anomalous results, the data are consistent with expectations derived from survey data and that the claim to have replicated the results can be supported. We also suggest that our experiences show that questions of replication in the social sciences cannot be resolved by statistical meta-analysis alone.