Following to some degree the 2010 AgMasters session on analyzing trial data, we will here take a closer look at results from comparisons designed to test the effects of crop inputs. At a basic level, we’ll consider the nature of certain inputs and what we may want to be able to say about such inputs before we have any of our own data. While statistical analyses can help us to draw conclusions, it’s usually more complicated than simply saying a product “works” or “doesn’t work.” We’ll think about how to set “significance levels” and how to deal with results that “hint” at a response but don’t show clearly that an input produces a response. The discussion will include presentation and analysis of some actual data on crop inputs from recent trials. The objective will be to develop a realistic approach to testing, using, and (possibly) marketing such products.