The first of these three models used a single indepen-

dent variable, stage. The regression was significant

F (5, 136) = 5.92, p = .0001. The squared multiple corre-

lation coefficient equaled .18.

 The second model used sex and stage as independent

variables. These findings are summarized in Table 18.


Table 18

Regression Analysis of Rectocolon


Cancer Cases:


Source df F P


Model 6 5.71 .0001

 Sex 1 4.03 .0467

 Stage 5 5.27 .0002

 2
 R = .2025



The regression was also significant at the .05 level. The

partial regression coefficients of both independent variables

were also significant at the .05 level.

 Model two, with sex and stage as independent variables,

was selected for use in predicting the survival of rectocolon


Model Two