This paper introduces a utilitarian confidence testing statistic called Risk Inclination Model (RIM) which indexes all possible confidence wagering combinations within the confines of a defined symmetrically point-balanced test environment. This paper presents the theoretical underpinnings, a formal derivation, a hypothetical application, and published results of using RIM during mathematics and science testing. This paper advocates the position that such a model provides researchers and policy makers a new statistic to measure an examinee's reflective confidence toward his/her knowledge during testing. Adding RIM to the reflection cluster of a global assessment framework like OCED/PISA could enhance future self-confidence assessments within knowledge-based societies. (Contains 2 tables and 1 figure.)