Correspondence Analysis of Risk.

Correspondence is determined by distance away from horizontal and vertical axis. The cloud of raw data points in the middle of the graph have no relationship to one another and therefore are close to the horizontal and vertical midpoints of the graph. The points are further away from the vertical midpoint but close on the horizontal axis, which indicates a good relationship between the predictor variables. However, the points far away from the vertical midpoint and far away from the horizontal midpoint indicates a poor relationship. Therefore participants' selections for lifetime breast cancer risk were consistent in their initial and repeated survey responses. Responses for ovarian cancer lifetime risk varied widely, indicating that the participants were guessing their response selections.