Federal law requires that state accountability systems include "another academic indicator" for elementary and middle schools in addition to academic achievement. Nearly all states use a measure of student longitudinal growth as their other academic indicator. But how should states decide which growth model to use? The Center for Assessment has helped several states choose or revise their growth models, and we find that these decisions involve much more than technical factors. From a strictly technical perspective, there isn't a "right" or "wrong" approach. To choose the right approach for their states, education leaders should consider their policy priorities, the intended uses of the growth results, any anticipated implementation challenges and, yes, technical factors. These things can be difficult to reconcile. This short paper suggests that states and districts use a principled approach to establish their most important values and walks leaders through key questions to ask. This process will help them evaluate potential growth models against an agreed-upon set of criteria.