Using the Discriminate function to classify the death and survivors cases of premature infants
journal of kerbala university,
2011, Volume 7, Issue 2, Pages 153-167
AbstractIt is known that the discriminate function is used to classify items into one or more groups depending on discriminate variables. Each group has characteristics that are available in its items and differs from the others. In order to investigate whether there are significant differences between two groups (the death of premature infant) in relation to the independent variables , to specify these independent variables that play a significant role of differing the categories of the followed variable, to divide the cases among the followed variable depending on the value of independent variables, and to evaluating the accuracy of division (as a percentage) i.e testing the theoretical side in the light of the classification whether the expected is true or not we use the discriminate function.
If the data that represent the death or survivors of the premature infant is the response variable, Sex (male, female), Hospital (kerbala, women), The way of delivery (normal, suzerain, other), The cause of admission (Least of weight, Shortage of pregnancy period, congenital pneumonias ,Meconium Aspiration syndrome ,Respiratory distress syndrome , Neonatal jaundice ,Hematological disorder of fetus , Septicemia , congenital anomalies ,Birth asphyxia ,Birth trauma ,other disorders) The age of the mother , the weight of the infant, the period of pregnancy as independent variables that affect on the baby’s staying alive.
Through out the research, we found that the best variables of distinction were the period of pregnancy and the cause of admission. The relation factor between the discriminate function and the period of pregnancy was 0.72., the relation factor between the discriminate function and the cause of admission was 0.511. It is also found that the highest binary relation factor was between the period of pregnancy and the weight of the infant which is 0.716, followed by the binary relation factor between the type of delivery and the weight of the baby which is 0.327. It is also found that there are differences of significant reference between the group of dead and alive babies in relation to the independent variables. That was clear through noting the difference between the origins of the discriminate function. The independent variable (period of pregnancy) had the biggest effect in increasing the distinction between the two groups. This is why the function called with its name (period of pregnancy function)
- Article View: 31
- PDF Download: 17