Robust and reliable quantitative proliferation assessment tools have the potential to contribute significantly to a strengthened nonproliferation regime and to the future deployment of nuclear fuel cycle technologies. Efforts to quantify proliferation resistance have thus far met with limited success due to the inherent subjectivity of the problem and interdependencies between attributes that lead to proliferation resistance. We suggest that these limitations flow substantially from weaknesses in the foundations of existing methodologies--the initial data inputs. In most existing methodologies, little consideration has been given to the utilization of varying types of inputs--particularly the mixing of subjective and objective data--or to identifying, understanding, and untangling relationships and dependencies between inputs. To address these concerns, a model set of inputs is suggested that could potentially be employed in multiple approaches. We present an input classification scheme and the initial results of testing for relationships between these inputs. We will discuss how classifying and testing the relationship between these inputs can help strengthen tools to assess the proliferation risk of nuclear fuel cycle processes, systems, and facilities.