LEARNING UNIT EIGHT

HYPOTHESIS TESTS – MEANS AND

PROPORTIONS
 GENERAL INTRODUCTION
We have now reached the point in this module where we start to
apply inferential statistics to information obtained from a sample,
in order to be able to draw conclusions about the population from
which the sample was selected.

Specifically, we will be testing whether a claim made regarding a

population parameter is reasonable, based on the information
obtained from an associated sample.

The process of testing the validity of a claim as to the true value of

any population parameter is referred to as hypothesis testing.

In this learning unit, we will be testing claims regarding the true

value of a population mean, or a population proportion.
 The Five Step Procedure For
Hypothesis Testing
Step 1: Formulating The Null And Alternative Hypotheses

• Null hypothesis: denoted as 𝑯𝟎 , this statement claims that

the population parameter is either equal to, less than or
equal to, or more than or equal to a particular value;

• Alternative hypothesis: denoted as 𝑯𝟏 , this statement

indicates that the value of the population parameter is the
exact opposite of that suggested by the null hypothesis, i.e.
the population parameter is either not equal to, greater
than, or less than a particular value.
The hypothesis test always tests the validity of the null
hypothesis, based on the data available from the
associated sample.

We say that the null hypothesis is rejected, if the

evidence obtained from the sample indicates that the
true value of the population parameter is significantly
different from that indicated by the null hypothesis.

We say that the null hypothesis is not rejected, if the

evidence obtained from the sample indicates that the
true value of the population parameter is not significantly
different from that indicated by the null hypothesis.
Step 2: Determine The Areas Of Acceptance And
Rejection, And Formulate The Decision Rule

The area of acceptance is that range of values for

the sample statistic that will lead to the null
hypothesis not being rejected.

The area (or areas) of rejection is that range of

values for the sample statistic that will lead to the
null hypothesis being rejected.

Examples are given on the following slides of the

areas of acceptance and rejection for each of the
test types.
Two-Tailed Test (𝑯𝟎 Indicates That The Population Parameter Is
Equal To A Specified Value)

Area Of 𝑯𝟎 : 𝝁 𝒐𝒓 𝝅 = 𝒌 Area Of
Rejection Area Of Acceptance Rejection
Upper Tailed Test (𝑯𝟎 Indicates That The Population Parameter
Is Less Than Or Equal To A Specified Value)

𝑯𝟎 : 𝝁 𝒐𝒓 𝝅 ≤ 𝒌 Area Of
Area Of Acceptance Rejection
Lower Tailed Test (𝑯𝟎 Indicates That The Population Parameter
Is Greater Than Or Equal To A Specified Value)

Area Of 𝑯𝟎 : 𝝁 𝒐𝒓 𝝅 ≥ 𝒌
Rejection Area Of Acceptance
Errors In Hypothesis Testing

• Type I error: rejecting a true null hypothesis;

• Type II error: failing to reject a false null
hypothesis

Selecting Significance Level (𝜶)

• 𝛼 represents the maximum probability that we

will make a Type I error;
• The most commonly used levels are 10%, 5% and
1% (0.10, 0.05 and 0.01).
If the null hypothesis uses the equals sign, “=“,
the test is two-tailed.

As we have seen on a previous slide, this type of

test has two areas of rejection, in the left-hand
and right-hand tails of the distribution curve.

The total area in the two areas of rejection is

equal to the significance level, 𝝈. The area in
each individual area of rejection is therefore
equal to 𝝈Τ𝟐.
If the null hypothesis utilises the greater than or equal to
sign, “≥”, then the test is referred to as a lower-tailed
test. There is only one area of rejection, in the lower-tail
of the distribution curve. The entire area in this single
area of rejection is equal to 𝜎.

If the null hypothesis utilises the less than or equal to

sign, “≤”, then the test is referred to as an upper-tailed
test. Again, there is only one area of rejection, this time
in the upper-tail of the distribution curve. The entire area
in this single area of rejection is equal to 𝜎.

Both of the above tests are referred to as single-tail or

one-tailed tests.
Identifying The Critical Value(s) For The Test, And
Formulating The Decision Rule

The type of test, and the level of significance, together

determine the critical value or values.

The critical values determine the areas of acceptance and

rejection under the distribution curve:

• A one-tailed test generates one critical value;

• A two-tailed test generates two critical values.

The critical values are then used to formulate the

decision rule – the condition or conditions under which
we will reject the null hypothesis.
For a test involving the 𝑧-distribution, we
denote a critical value as 𝒛𝒄𝒓𝒊𝒕 and the
sample statistic as 𝒛𝒄𝒂𝒍𝒄 .

The following decision rules are then

applied:

• Two-Tailed Test: We reject 𝐻0 if:

𝒛𝒄𝒂𝒍𝒄 < −𝒛𝒄𝒓𝒊𝒕 or 𝒛𝒄𝒂𝒍𝒄 > +𝒛𝒄𝒓𝒊𝒕

• Upper-Tailed Test: We reject 𝐻0 if:

𝒛𝒄𝒂𝒍𝒄 > +𝒛𝒄𝒓𝒊𝒕

• Lower-Tailed Test: We reject 𝐻0 if:

𝒛𝒄𝒂𝒍𝒄 < −𝒛𝒄𝒓𝒊𝒕

Step 3: Compute The Corresponding Sample
Statistic

The sample statistic is computed based on the

information obtained from the sample.

The formula to compute the sample statistic will be

dependent upon the distribution associated with
the nature of the test being performed.

For this learning unit, the 𝑧 and the 𝑡 distributions

will be encountered.
Step 4: Compare The Sample Statistic Value With
The Decision Rule

• If the sample statistic value falls into the area(s)

of rejection, 𝐻0 is rejected;
• If the sample statistic value does not fall into the
area(s) of rejection, 𝐻0 is not rejected.

Step 5: Draw Statistical & Management Conclusion

Indicate whether the 𝐻0 is being rejected or not at

the relevant significance level, and indicate the
implication for the claim that we were asked to test.
 Hypothesis Test For
Single Population Mean
Sample Statistic Formula If 𝝈 Is Known

𝑥ഥ − 𝜇
𝑧𝑐𝑎𝑙𝑐 = 𝜎
𝑛

Sample Statistic Formula If 𝝈 Is Unknown And 𝒏 ≤ 𝟒𝟎

𝑥ഥ − 𝜇
𝑡𝑐𝑎𝑙𝑐 = 𝑠
𝑛
Sample Statistic Formula If 𝝈 Is Unknown And
𝒏 > 𝟒𝟎

𝑥ഥ − 𝜇
𝑧𝑐𝑎𝑙𝑐 = 𝑠
𝑛
Summary: When To Use z And t When 𝝈 Is
Unknown

• If the population standard deviation is

unknown and the sample size is small (n ≤ 40),
use t;

• If the population standard deviation is

unknown and the sample size is large (n > 40),
use z.
 Hypothesis Test For
Single Population Proportion
Sample Statistic Formula If 𝒏 > 𝟒𝟎

𝑝− 𝜋
𝑧𝑐𝑎𝑙𝑐 =
𝜋(1 − 𝜋)
𝑛