CONTENTS 



Page 



INTRODUCTION 1 



APPLICATIONS OF MODIFIED PROGRAM 3 



Mortality 3 



Cull Estimation 4 



Canker Inactivation 4 



SAMPLING ASSUMPTIONS OF ALGORITHM 5 



WALKER-DUNCAN ALGORITHM 6 



MODIFICATION OF ALGORITHM 8 



STATISTICAL CRITERIA FOR GOODNESS-OF-FIT 10 



GUIDES FOR USE OF RISK 12 



Control Cards 12 



Data Preparation 15 



Program Output 16 



Error Diagnostics 16 



DISCUSSION AND CONCLUSIONS 18 



LITERATURE CITED 18 



ABSTRACT 



Describes an algorithm for fitting the relationship between a 

 dichotomous dependent variable and a number of independent var- 

 iables. The specific functional form utilized is the logistic func- 

 tion. The algorithm, proposed and programed by Walker and 

 Duncan (1967), has been generalized by adding a subroutine to 

 handle data transformations and by introducing procedures to 

 handle data sampled with unequal probability. A user's guide 

 discusses makeup of program control cards, data transfor- 

 mation capabilities, program output, and program-generated 

 errors. 



