Breast Cancer

 The development of a regression equation for the pre-

diction of the survival of breast cancer patients also

initially involved exploratory regression analysis of two

models, one with the main effects of independent variables

and one with main and interaction effects. Both regression

analyses were based on 96 cases of the original sample of

101. The cases excluded were those with missing values for

histology and/or stage. As for the lung sample, a comparison

of the distribution of values on all variables revealed only

slight differences between the two groups. A comparison of

the group of 96 to the sample of 101 is given in Tables

42-46, Appendix F.

 The first analysis used age, race, stage, histology,

initial treatment, and subsequent treatment as the indepen-

dent variables. The dependent variable was survival; that

is, the number of months between the diagnosis and death.

The regression was significant with F (36,59) = 2.28, p =

.0024 and R2 = .58.

 The predictive efficacy of each independent variable

was evaluated through partial regression analysis, the

results of which are summarized in Table 15.