Study Guide: Activity 2.2: Question 1 on page 51.

How to calculate the p-value:

The calculation of 𝑝𝑝 − value was not shown in the study guide of STA1502 because

this concept is demonstrated in STA1501 in Chapter 11: Hypothesis testing of the

mean and for proportion.

The 𝑝𝑝 −value helps us to draw a conclusion from a testament made. E.g. During the

registration period at UNISA, a student might state that he/she will pass the module

with 60%. This is a statement that can be true or not. The statement made can be

called hypothesis. Since it is a hypothesis, we must test it so that we can sure if we

can accept the statement or reject it.

In practice, the statement will be made about a population mean or a population

proportion then the hypothesis will be called hypothesis of the mean or hypothesis for

the proportion. E.g. The average production of a company of bottle of water is 3000

per day, this is a statement. Here the null hypothesis 𝐻𝐻0 : 𝜇𝜇 = 3000 against the

alternative hypothesis 𝐻𝐻1 : 𝜇𝜇 = 3000.

STA1501 is dealing with one population at the time and the population mean is

denoted by 𝜇𝜇 as shown in the above example.

In STA1502, we are dealing with two populations at the same time. That why the aim

is to compare the two population means if they are different or not. The two-population

means are denoted 𝜇𝜇1 and 𝜇𝜇2 , respectively. The same way we will compare two

population proportions.

In comparing two population means, there are two options: either the two populations

are equal, or the two population are not equal. Furthermore, when the two populations
are not equal, that means one population is greater than the other one. Let us illustrate

this concept by describing it in the language of the hypothesis by saying:

• The null hypothesis 𝐻𝐻0 : The two-population means are equal. If we use the

mathematical notations, we will say 𝐻𝐻0 : 𝜇𝜇1 = 𝜇𝜇2 . This is the same as 𝐻𝐻0 : 𝜇𝜇1 −

𝜇𝜇2 = 0.

• The alternative hypothesis 𝐻𝐻1 : The two-population means are not equal. If we

use the mathematical notations, we will say 𝐻𝐻1 : 𝜇𝜇1 ≠ 𝜇𝜇2 . This is the same as

𝐻𝐻1 : 𝜇𝜇1 − 𝜇𝜇2 ≠ 0.

=: This sign is called “equal to”.

≠ : This sign is called “not equal to”.

Let us go to the question of the activity 2.2 of the study on page 51: This is the

difference between two population proportions. To do this, let us replace 𝜇𝜇1 with 𝑝𝑝1 and

𝜇𝜇2 with 𝑝𝑝2 .

The question is to test if the 𝑯𝑯𝟎𝟎 : 𝒑𝒑𝟏𝟏 = 𝒑𝒑𝟐𝟐 against 𝑯𝑯𝟏𝟏 : 𝒑𝒑𝟏𝟏 ≠ 𝒑𝒑𝟐𝟐 at the 10% level of

significance.

You can agree with me that this is a test to determine whether two population

proportions differ. This is because we can all see 𝑯𝑯𝟏𝟏 : 𝒑𝒑𝟏𝟏 ≠ 𝒑𝒑𝟐𝟐 , a clear indication of two

population proportions differs. This is called also a two-tailed test for using the sign ≠.

Let us go now calculate the p-value.

𝑃𝑃 −value will assist to draw a conclusion if the null hypothesis 𝐻𝐻0 is rejected or not.

𝑃𝑃 −value is the smallest area that is called the rejection region when using a normal

distribution.
To draw the conclusion, we will use a rule: when p-value is less than the level of

significance like 0.05 for instance, we can conclude that the null hypothesis 𝑯𝑯𝟎𝟎

is rejected in favour of the alternative hypothesis 𝑯𝑯𝟏𝟏 .

To calculate the 𝑝𝑝 −value, we need a test statistic value otherwise it is difficult to

calculate it. From the solution of the activity 2.2 on page 53 of the study guide, you

can see that we started by calculating the test statistic for proportion so that we can

be able to calculate the 𝑝𝑝 −value. The value of the test statistic 𝒁𝒁 = −𝟏𝟏. 𝟖𝟖𝟖𝟖𝟖𝟖𝟖𝟖.

To calculate the 𝑝𝑝 −value using the test statistic: this is written 𝑃𝑃(𝑍𝑍 < −1.8293)

because the value of 𝑍𝑍 is negative. If 𝑍𝑍 was positive it should be 𝑃𝑃(𝑍𝑍 > 1.8293).

Let us show how to calculate the 𝑝𝑝 −value: 𝑃𝑃(𝑍𝑍 < −1.8293) = 𝑃𝑃(𝑍𝑍 < −1.83), We

need 2 decimals because we will use the standardized normal tables called Table 3

from the textbook or Table 1 from the study guide (see page 6 and 7).

To calculate the 𝑝𝑝 −value of 𝑃𝑃(𝑍𝑍 < −𝟏𝟏. 𝟖𝟖𝟖𝟖), let us write 𝟏𝟏. 𝟖𝟖𝟖𝟖 = 𝟏𝟏. 𝟖𝟖 + 𝟎𝟎. 𝟎𝟎𝟎𝟎.

Let me draw an illustration of the Table 1, to show how we read the values.

Table 1

𝑍𝑍 0.00 0.01 0.02 0.03 … 0.99

-3.0

-2.9

-1.8 0.0359 0.0351 0.0344 0.0336 … 0.0294

… …

-0.0 0.5000
The 𝑝𝑝 −value = 𝑃𝑃(𝑍𝑍 < −1.83) = 0.0336, because from the hypothesis, we said it was

a two-tailed test therefore the 𝑝𝑝 −value is multiplied by two.

The 𝑝𝑝 −value = 2 × 𝑃𝑃(𝑍𝑍 < −1.83) = 2 × 0.0336 = 0.0672.

Since 𝑝𝑝 −value helps us to draw the conclusion, now let us use the rule: reject 𝑯𝑯𝟎𝟎 if

𝒑𝒑 −value is less than 0.10. Since 0.0672 is greater than 0.10, we can conclude that

𝐻𝐻0 is not reject at 10% level of significance.

Remark

Since 𝑝𝑝 −value is the smallest area that we can reject the null hypothesis. We will

always use the Table 1 with the small shade to calculate the 𝑝𝑝 −value. If the test

statistic is positive, we will still use the same Table 1 with a small shade because a

normal distribution is symmetrical. That means, the shaded portion at the negative

side is equal to the corresponding shaded area at the positive side.