1. Imagine you have the following data: Stratum-specific Stratum-specific Stratum-specific estimate 1 estimate 2 estimate 3 Crude estimate OR=6.25 OR=1.24 OR=0.23 OR=2.75 This best represents the scenario of: O Confounding, but no effect modification O No confounding and no effect modification O Effect modification, but no confounding O Effect modification, so confounding is irrelevant O Measurement error