Document Type : Research Paper
Author
Anbar University - College of Education for Pure Sciences
Abstract
In this paper ,I have studied a new(which cum.f approximate method)is used for calculating
the approximate normal boundaries by using the Neyman allocation in case of studying classes that have
heterogeneous samples and compared it with the most efficient approximate method ,which is ) (cum.f 6
5
.It
turned out through this study that the new method is more efficient than the old method in case of the
uniform distribution and for all values of d ,while in case of the normal distribution , the efficiency of the
new method on the old one depends on value (σ ) ,if it was (σ < 0.2287 ) the old method would be more
efficient than the new one ,while if it was (σ = 0.2287 ) ,the two methods would be at the same level of
efficiency ,while if it was (σ >0.2287 ) ,the new method would be more efficient than the old one ,As for the
exponential distribution ,the new method would not always be efficient but during specific moment
depending on value (λ ) ,while if it was (λ < 3.042 ) ,the new method would be more efficient than the old
one ,if it was (λ = 3.42 ) the two methods would be at the same level of efficiency , while if it was ( λ >
3.042) the old method would be more efficient than the new one.
)
