Page 57 - Basic Statistics

52

In the hypothesis testing are two kinds of errors, namely the type I error

and the type II error. Type I error is a mistake to reject H0 when in fact H0 is
true. Type II error is the error accept H0 when in fact H1 is true.

Probability of a type I error occurs is denoted by  (alpha), which is often

also called the significance level. While probability of type II error occurs

denoted by  (beta). In statistical hypothesis testing, probability of the two types

of errors are expected balanced and tailored to the problem under study.
Probability value 1 -  is called the confidence interval, expressed probability

accept H0 when it is true. Probability value 1 -  is called the power of the test,

states the probability to reject H0 when H0 is false.

4.1 HYPOTHESIS TESTING FOR SINGLE
POPULATION MEAN

The data of a sample taken from a population can be used to assess the

characteristics of its population. One of the most interesting characteristics of

the population to be studied is the mean value. There are three forms of the

formulation of a test hypothesis for the mean value of single population,

namely:
1 H0 :  = o
H :   o
1

2 H0 :   o
H :  < o ( H1 )
1

3 H0 :   o
H :  > o
1

In the conduct of any hypothesis test, there are five steps to be followed :
a. Formulating a test hypothesis, namely the null hypothesis (H0) and the

alternative hypothesis (H1). It is one of the formulas above.

b. Calculate the sample statistics, ie the mean value of the sample X

~~* CHAPTER 4 HYPOTHESIS TESTING *~~