This chapter has dealt with some complex extensions of previously discussed designs that are intended to provide either quick, approximate information about the differential effects of the components...

This chapter has dealt with some complex extensions of previously discussed designs that are intended to provide either quick, approximate information about the differential effects of the components of intervention or logically sound information on the same topic using an elaborate design. The alternating intervention design, A-B/C-(B or C) begins from a baseline, followed by an intervention phase in which two or more interventions are randomly alternated, possibly with more than one practitioner involved. Based on your evaluation of how each intervention went, you try the “best” one and evaluate it in the regular way. This is a fast and economic approach to looking at multiple interventions to determine, approximately, the most effective one. The interaction design is a very powerful logical design, but a cumbersome one as well. The A-B-A-B-BC-B-BC design is the minimum arrangement for full logical determination of the additive and separate effects of multiple interventions, and it illustrates many of the logical principles of design we have been discussing in this book. There are other complex designs, but these essentially have been used mainly in laboratory settings and not in field settings. However, the important point is that you should feel free to be creative in constructing a design as long as your interpretation of results follows the basic principles of the design that we have described.