This article argues with a literature review that a simplistic distinction between strong and weak evidence hinged on the use of randomized controlled trials (RCTs), the federal "gold standard" for generating rigorous evidence on social programs and policies, is not tenable with evaluative studies of complex, field interventions such as those found in education. It introduces instead the concept of grades of evidence, illustrating how the choice of research designs coupled with the rigor with which they can be executed under field conditions, affects evidence quality progressively. It argues that evidence from effectiveness research should be graded on different design dimensions, accounting for conceptualization and execution aspects of a study. Well-implemented, phased designs using multiple research methods carry the highest potential to yield the best grade of evidence on effects of complex, field interventions.