ABSTRACT
This chapter identifies the appropriate algebraic structures for the study of refraction power matrices and their statistical summaries, following with reference to data from the Oxford County (Ontario, Canada) Vision Screening Program and the Visual Acuity Study Sample. It was also observed that under certain projective restrictions the refraction matrices are homeomorphic to the open solid torus D × S and that such mapping has many potential data-analytic consequences. In particular, the torus renders itself as a summary space for the homogeneous (projective) data. Perhaps not surprisingly, but certainly reassuring, is the fact that the projective condition allowing for such homeomorphism is simply that the principal axes must be aligned with the canonical axes of the plane.
