Lesson REJECTION REGION AND

2 LEVEL OF SIGNIFICANCE
In this lesson, you are going to learn about test direction, rejection region, and
level of significance which are very important in making decisions about the claim or
hypothesis. In order to have a clear idea about these things, read thoroughly the texts
that follow:

Learning Concepts

Recall that the normal curve evolved from the probability distribution.
With the area under the curve being equal to 1, it has become a
mathematical model in hypothesis testing. The areas are probability values that we
need for decision – making. In hypothesis testing we determine the probability of
obtaining the sample results if the null hypothesis is true. Thus, the calculations can
be graphically represented using the normal curve. The greater than (>) the mean
direction can be shown at the right tail of the curve just as the less than (<) the mean
direction can be shown at the left tail (Belecina, et al.,2016)

Below, are the graphical representations of the non – directional and

directional tests.

Non – directional
(Two – tailed)

Directional
(One – tailed, right tail) 𝛼𝛼

Directional
(One – tailed, left tail) 𝛼𝛼
1 − 𝛼𝛼

16
 To determine whether the test is directional or non-directional we simply
examine the alternative hypothesis. If it uses the “not equal to” symbol (≠), then the
test is non-directional but if it utilizes the “less than (˂)“ or “greater than (˃)” symbol,
the test is directional.
The following simple steps are suggested in order to determine the
direction of the test.
1. State the null and alternative hypotheses.
2. Examine the symbol used in the alternative hypothesis.
3. Make a decision. (Directional or Non-directional)

Here are some keywords that denote direction:

Going to the right (right – tailed) Going to the left (left – tailed)

Greater efficient lesser fewer

Increases improves decreases ineffective
Augment effective diminish
Advances more delayed

Now, let us examine an example of a directional and non – directional test.

Example 1: A baker claims that the cake he developed is fortified with vitamin A.
Step 1: The hypotheses are;
Ho : The developed bread is not different with other bread.
Ha : The developed bread is fortified with vitamin B.
Step 2: The alternative hypothesis utilizes the ˃ symbol because of the
word fortified which means “to strengthen” or “to add”
Step 3: The test is directional or one-tailed (right-tailed).

Example 2: A combination of fruits provides the minimum daily requirement for vitamin
C.
Step 1: The hypotheses are;
Ho : The amount of vitamin C is equal to the daily requirement.
Ha : The amount of vitamin C is not equal to the daily requirement.
Step 2: The alternative hypothesis utilizes the ≠ symbol.
Step 3: The test is non-directional or two-tailed.

In making decisions, conclusions are formed and these conclusions are the bases
of actions. But this is not always the case in Statistics because the decisions that are
made were based on sample information. The best thing to do is to control the
probability with which error occurs.

17
 The probability of committing a Type I error is denoted by the Greek letter α (alpha)
while the probability of committing a Type II error is denoted by β (beta). These
probabilities are shown on the table below.

Types of Errors and Their Probability

Error in decision Type Probability Correct decision Type Probability

Reject a true Ho I α Accept a true Ho A 1–α
Accept a false Ho II β Reject a false Ho B 1–β

Errors can be controlled by assigning small probability values to each of them.

The most commonly used probability values for α and β are 0.05 and 0.01. The
probability assigned to each depends on its seriousness. The more serious the errors,
the less willing we are to have it occur. So, a smaller probability will be assigned. As
seen from the table above, 1 – α is the probability of a correct decision when the null
hypothesis is true, and 1 – β is the probability of a correct decision when the null
hypothesis is false. 1 – β is called the power of the statistical test since it is the measure
of the ability of a hypothesis test to reject a false null hypothesis which is very
important. (McClave & Sincich, 2003)

Below is the graphical representation of the decision errors under the normal curve.

α
Reject Ho Reject Ho Reject Ho

Notice that the rejection region of a directional test is in one tail while the
non – directional test the rejection region is distributed to the two tails of the
curve.
A Rejection region refers to the region under the normal curve where the
value of the test statistic lies for which the null hypothesis will be rejected. This
region is sometimes called critical region.
Therefore, if the computed statistic lies on the rejection region, then we
reject the null hypothesis. But, if it is found outside the rejection region, we
do not reject (accept) the null hypothesis.
Notice also that there is a line that separates the rejection region from the
non-rejection region denoted by 1 – α. This line passes through the
confidence coefficients, which are also called critical values.

18
 The table below shows the critical values for the z – statistic.

Significance Level ( α ) 0.10 0.05 0.03 0.02 0.01

Confidence Level ( 1 – α ) 90% 95% 97% 98% 99%
One – tailed (Directional) 1.28 1.64 1.88 2.05 2.33
Two – tailed (Non-directional) 1.64 1.96 2.17 2.33 2.58

The table of critical values of z shows that for a 95% confidence level, the critical
values for a non–directional test are –1.96 and +1.96. The negative ( - ) and positive
(+) signs tell that the rejection region is found on both tails of the distribution.
Moreover, for a directional test the critical value at 95% confidence level is +1.64 for
right–tailed and –1.64 for left – tailed.

To understand better the table, study the succeeding examples.

Example 1. Is the computed z = 2.0 at 90% confidence level, two-tailed found in the
rejection region or acceptance region.

Steps Solution
Step 1: Determine the critical From the table, the critical value of z is ±1.64
value of z at 90% the ± sign is added because the direction is
confidence level, two- two-tailed.
tailed. Significance Level (α) 0.10 0.05 0.03 0.02 0.01
Confidence Level (1 –
90% 95% 97% 98% 99%
α)
One – tailed 1.28 1.64 1.88 2.05 2.33
Two – tailed 1.64 1.96 2.17 2.33 2.58

Step 2: Compare the given z computed z critical

computed value of z 2.0 ˃ 1.64
with the critical value.
The computed value of z is greater than
the critical value from the table.
Step 3: Location in the normal Location
curve of z=2.0

Accept Ho
Reject Ho Reject Ho

─1.64 +1.64

The computed z lies in the rejection

region because under the normal curve
z=2.0 is at the right of 1.64.

19
Example 2. At what region is the computed z = 1.33 at 1% significance level, one-
tailed located?
Steps Solution
Step 1: Determine the critical From the table, the critical value of z is 2.33
value of z at 1% (0.01 the ± sign is omitted because the direction is
in decimal form) one-tailed.
significance level, one- Significance Level (α) 0.10 0.05 0.03 0.02 0.01
tailed. Confidence Level (1 –
90% 95% 97% 98% 99%
α)
One – tailed 1.28 1.64 1.88 2.05 2.33
Two – tailed 1.64 1.96 2.17 2.33 2.58

Step 2: Compare the given z computed z critical

computed value of z 1.33 ˂ 2.33
with the critical value.
The computed value of z is less than the
critical value from the table.
Step 3: Location in the normal Location
curve of z=1.33

Accept Ho α
Reject Ho

2.33

The computed z lies in the acceptance

region because under the normal curve
z=1.33 is at the left of 2.33

Key Concepts:

A non – directional test is also called a two–tailed test.

A directional test may either be right–tailed or left–tailed.
1 – α is the probability of a correct decision when the null hypothesis is true.
1 – β is the probability of a correct decision when the null hypothesis is false. It is
the power of the statistical test.
Rejection region refers to the region under the normal curve where the value
of the test statistic lies for which the null hypothesis will be rejected.
Acceptance region refers to the region under the normal curve where the
value of the test statistic lies for which the null hypothesis will not be
rejected.

20