Page 64 - Basic Statistics

59

4.2 HYPOTHESIS TESTING
FOR TWO POPULATION MEANS

A research often aims to compare the average value of two independent

populations. These two populations are parent populations from which two
samples are drawn. Statistical method provides several ways to test differences

in the average value of two populations. In parametric statistical method, the

test can be carried out differences with the approach normal distribution

(standard normal test-z), and the approach student-t distribution (Student's t-

test). Both approaches this distribution can be done if the data is quantitative
and continuous, so that the measurement data is feasible to apply the

assumption of a normal distribution.

4.2.1 MEAN DIFFERENCE TEST USING STANDARD NORMAL TEST

Approach a normal distribution; The standard normal test-Z used if
variance of the two populations are known or sample variance ( S ) has
2
stabilized to estimate population variance (  ). Sample variance ( S ) has
2
2
stabilized to estimate population variance  when the sample size n  30.
2
Similarly, for sample size n  30, mean ( X ) of the random variable X (unknown

distribution) will form a normally distribution.

Hypothesis test to test the equality of the average of two independent
populations (two-tailed test) is formulated as follows:

H0 : μ = μ
1
2
( N1)
H1 : μ  μ
1
2

Where,

μ = Mean of the first population
1
μ = Mean of the second population
2

~~* CHAPTER 4 TESTING HYPOTHESIS *~~