Twelve SQUEACs would be time-consuming. You can take a sample of UCs. The approach that I use is to build a tree or hierarchy of fixed and program effects with UCs as "leaves" at the end nodes of the tree. Here is an example: +-- URBAN -- UC1, UC2, UC3 | | District --+ | +-- NGO supported -- UC4, UC5, UC6 | | +-- RURAL --+ | +--- NOT upported -- UC7 .. UC12 In this example we have three end nodes (this is the minimum number of SQUEACs needed). You might pick a UC at random from each of the end-nodes. It is a good idea to start by listing all factors that you think might have a big effect program operation / coverage and build a bushy tree which you can then "prune" back to include only the most likely major factors. Be sure that the tree can acomodate all UCs. I hope this helps.