We have a project within our sight that encourages us to split data collection between CAWI and CAPI (important part of the population consists of people who don't use Internet, but at the same time we have to meet our deadlines and cost limits). Have you ever tried such a merger? Any reflections about mode effects (scale factor, different heuristics)? Before we field a multimode CBC survey we have to be sure it will blend painlessly. I thought about some experimental setting which could help to elicit mode effects, but it seams to be hard to make it right (bugget).