Deciding what data to collect for an experiment is a critical decision that has substantial consequences for how resources are spent, what information can be gleaned from the results of the experiment, and what conclusions can be drawn.
Knowing what you want from an experiment and what will constitute success are key to making good decisions, as often there may be multiple goals that are important. Likewise, successful experimentation is about managing your budget and your time well, so being able to examine and compare performance against the cost of the experiment should be done explicitly across a range of budget possibilities.
In designing hundreds of experiments throughout her career as a consultant and collaborator at Virginia Tech and Los Alamos National Laboratory, Chirstine Anderson-Cook has learned (sometimes the hard way) the importance of clarifying the goal(s) of the experiment with the scientists and engineers and then comparing alternatives to find the designed experiment that best matches the objectives.
The good news is this has become dramatically easier with JMP® 17 and the introduction of Design Explorer, which provides a simple way to construct and contrast multiple designs. JMP has long been a world leader in generating excellent designs for a wide variety of purposes, but now it is even easier to generate multiple desirable choices with a single series of dialog boxes that can be compared and evaluated.
In this white paper, you'll learn best practices for looking more broadly at mulitple options for your experiments and how to use Design Explorer, a new feature in JMP 17, to generate a suitable suite of candidate designs to choose from.