Because of those percentages, for statistical power we dichotomized our sample into white patients/patients of color, with clear caveats about the limitations regarding generalizability, applicability to different populations, etc.

Comments