LECTURE 2

LIKELIHOOD RATIO TESTS

 The likelihood ratio (LR) test is a test of hypothesis in which two

different maximum likelihood estimates of a parameter are compared in order to
decide whether to reject or not to reject a restriction on the parameter.
 The critical region (and test statistic) for the test are determined by setting an
upper bound on the ratio (𝑚𝑎𝑥 𝐿0 / max 𝐿 ), where 𝑚𝑎𝑥 𝐿0 is the maximum value
of the likelihood 𝐿 under the restrictions imposed by the null hypothesis, and
𝑚𝑎𝑥 𝐿 is the overall maximum value of L for all allowable values of all parameters
involved.
 In the most common case when 𝐻0 and 𝐻1 together cover all possible values for the
parameters, this generalised test rejects 𝐻0 if:
max(likelihood under H0 )
< crtical value
max(Likelihood under H0 + H1 )
 Important results include the case of sampling from a 𝑁(𝜇, 𝜎 2 ) distribution. The
method leads to the test statistic:
𝑋̅ − 𝜇0
~𝑡𝑛−1 𝑢𝑛𝑑𝑒𝑟 𝐻0 : 𝜇 = 𝜇0
𝑆⁄√𝑛
 For test on the value of the mean 𝜇
 This is done based on the assumption that 𝜎 2 is unknown. If it is known, then the
Z-test is the “best” test.
 The method also leads to the test statistic:
(𝑛 − 1)𝑆 2 2
~𝜒𝑛−1 𝑢𝑛𝑑𝑒𝑟 𝐻0 : 𝜎 2 = 𝜎02
𝜎02

 For tests on the value of the variance 𝜎 2

Introduction to statistical tests

 A statistical test, is a method used to determine whether there is enough evidence
in a sample of data to infer that a certain hypothesis about a population parameter
is likely true.
 Statistical tests can be categorized as
(i) Left-tailed
(ii) Right-tailed
(iii) Two-tailed
 A statistical test is;
(i) a left- tailed if 𝐻1 states that the parameter is less than the value claimed in
𝐻0
(ii) a right-tailed if 𝐻1 states that the parameter is greater than the value
claimed 𝐻0

Page 1 of 19
 LECTURE 2

(iii) two-tailed if 𝐻1 states that the parameter is different (or not equal to) the
value claimed in 𝐻0

Types of Hypotheses
 A hypothesis is a statement about the value of an unknown parameter in the model.
 There are two types of hypotheses. Namely;
(i) Null hypothesis (𝑯𝟎 )
 This can be thought of as the implied hypothesis. “Null” meaning
“nothing.”
 This hypothesis states that there is no difference between groups or no
relationship between variables
 Usually the null hypothesis represents a statement of “no effect”, “no
difference” or , put another way, “ things haven’t changed”
 The null hypothesis is a presumption of status quo or no change.
 It is the initial claim about the value of a parameter.
 Always contain s some variation of equality as part of the relationship

(ii) Alternative hypothesis (𝑯𝒂 )

 This is a statement you will adopt in the situation in which the evidence
(data) is so strong that you reject 𝑯𝟎
 This hypothesis should state what you expect the data to show
 The alternative hypothesis contains the opposite inequality.
Remark;
 In statistical testing, the null hypothesis 𝐻0 always contains the equal symbol.
However in the null hypothesis, some statistical packages and texts also include the
inequality symbol that is opposite to the one shown in the alternate hypothesis.
 The mathematical construction of a statistical test uses the null hypothesis to
assign a specific number (rather than a range of number) to a parameter of interest
say 𝜇 for example.
 The null hypothesis establishes a single fixed value say 𝜇 so that we work with a
single distribution having a specific mean

Hypothesis testing procedure

 The standard approach to carrying out a statistical test involves the following steps
(i) Specify the hypothesis to be tested/state the hypotheses in terms of
population parameters
𝐻𝑜 : Null hypothesis, usually is the opposite of our research hypothesis. The
null hypothesis always includes equality.
𝐻𝑎 : Alternative hypothesis, corresponds to our research hypothesis. Does
not include equality.

Page 2 of 19
 LECTURE 2

𝐻0 ≥ ≤ =
𝐻𝑎 < > ≠

Make sure you match the signs, so they

are opposite of each other, unless your
professor ants 𝐻0 to always have a " = "

Example;
𝐻𝑜: The mean number of GVSU students enrolled in STA215 during
WINTER 2018 who speak English as a second language is 15
𝐻𝑎: The mean number of GVSU students enrolled in STA215 during
WINTER 2018 who speak English as a second language is not equal to 15
(ii) Calculate the Test Statistic
We use a corresponding sample statistics from a simple random sample to
challenge the statement made in 𝐻0 . We convert the sample statistic to a
corresponding value of the appropriate sampling distribution
(iii) Calculate the P-value
We use the sampling distribution of the test statistic and type of test to
compute the P-value of this statistic. Under the assumption that the null
hypothesis is true, the p-value is the probability of getting a sample statistic
as extreme as or more extreme than the observed statistic form our random
sample.
(iv) State the Conclusion
We conclude the test. If the P-value is very small, we have evidence to reject
𝐻0 and adopt 𝐻1 . What do we mean by “very small?’ we compare the p-vale
to the preset level of significance 𝛼. If the 𝑃 − 𝑣𝑎𝑙𝑢𝑒 ≤ 𝛼, then we say that
we have evidence to reject 𝐻0 and adopt 𝐻1 . Otherwise, we say that the
sample evidence is insufficient to reject 𝐻0
(v) Finally, we interpret the results in the context of application

Basic tests – single samples

 Under this subheading we are going to tackle the following;
(a) Testing a claim about a single mean(𝜇)
(i) Testing claims about a population mean (𝜇) when (𝜎) is known
(ii) Testing claims about a population the mean (𝜇) when (𝜎) is
unknown
(b) A testing claim about a single proportion(𝑝)
(c) Testing a claim a standard deviation or variance(𝜎 2 )

(a) Testing a claim about a single mean

Page 3 of 19
 LECTURE 2

 Knowing the sampling distribution of the sample test statistic is an essential

part of the hypothesis testing process.
 For tests of 𝜇, we use one of the sampling distributions of 𝑥̅ . That is, the
standard normal distribution or student’s t distribution
 The appropriate distribution depends upon our knowledge of the
population standard deviation 𝝈, the nature of the 𝒙 distribution, and the
sample size

(i) Testing claims about a population mean (𝝁) when (𝝈) is known
 In most real –world situations, 𝜎, is simply not known however, in some
cases a preliminary study or other information can be used to get a
realistic and accurate value for 𝜎

Requirements
 Let 𝑥 be a random variable appropriate to your application.
 Obtain a simple random variable (of size n) of 𝑥 values from which you
compute the sample mean 𝑥̅ . The value of 𝜎 is already known (perhaps
from a previous study) if you can assume that 𝑥 ha s anormal
distribution, then any sample size 𝑛 will work.
 If you cannot assume this, then use a sample size 𝑛 ≥ 30
In other words;
 The sample is a simple random sample
 Either or both of these conditions are satisfied; the population is
normally distributed or 𝑛 ≥ 30
Procedure of testing
(i) In the context of the application, state the null and alternative
hypotheses and set the level of significance 𝛼.
(ii) Use the known 𝜎, the sample size 𝑛 , the value of 𝑥 from the sample
and 𝜇 from the null hypothesis to compute the standardized sample
test statistic
𝑥̅ − 𝜇
𝑍=
𝜎/√𝑛
(iii) Use the standard normal distribution and the type of test, one-tailed
or two tailed, to find the p-value corresponding to the test statistic
(iv) Conclude the test. If 𝑝 − 𝑣𝑎𝑙𝑢𝑒 ≤ 𝛼, then reject 𝐻0 . If
𝑝 − 𝑣𝑎𝑙𝑢𝑒 > 𝛼 then fail to reject 𝐻0
(v) Interpret your conclusion in the context of the application

Example1

Page 4 of 19
 LECTURE 2

Sunspot have been observed for many centuries. Records of sunspots from ancient
Persian and Chinese astronomers go back thousands of years. Some archeologists
think sunspots activity may somehow be related to prolonged periods of droughts
in the south western United States.
Let 𝑥 be a random variable representing the average number of sunspots observed
in a
4-week period. A random sample of 40 such periods from Spanish colonial times
gave the following data (reference: M. waldmeir, sunspot Activity, international
Astronomical union Bulletin)
12.5 14.1 37.6 48.3 67.3 70.0 43.8 56.5 59.7 24.0
12.0 27.4 53.5 73.9 104.0 54.6 4.4 177.3 70.1 54.0
28.0 13.0 6.5 134.7 114.0 72.7 81.2 24.1 20.4 13.3
9.4 25.7 47.8 50.0 45.3 61.0 39.0 12.0 7.2 11.3

The sample mean is 𝑥̅ = 47.0. Previous studies of sunspot activity during this
period indicates that 𝜎 = 35. It is thought that for thousands of years, the mean
number of sunspots per 4-week period was about 𝜇 = 41.
Sunspot activity above this level may (may not) be linked to gradual change. Do the
data indicate that the mean sunspot activity during the Spanish colonial period was
higher than 41? Use 𝛼 = 0.05

Solution

(i) Establish the null and alternative hypotheses

To determine the type of the test, identify the alternative hypothesis 𝐻1 .
Note that the claim to be tested is 𝜇 > 41
Note that the symbolic form that must be true when the original claim is
false is 𝜇 ≤ 41
Now identify the null and alternative hypotheses. The null hypothesis
(denoted by 𝐻0 ) is a statement that the value of a population parameter
(such as proportion, mean or standard deviation) is equal to some claimed
value.
The alternative hypothesis (denoted by 𝐻1 ) is a statement that the
parameter has a value that somehow differs from the null hypothesis
Consider the two symbolic expressions obtained so far , 𝜇 > 41 and 𝜇 ≤ 41.
The alternative hypothesis 𝐻1 is the one not containing equality. The null
hypothesis 𝐻0 is the symbolic expression that the parameter equals the
fixed value being considered. The null hypothesis and alternative
hypothesis are;
𝐻0 : 𝜇 = 41 𝑣𝑠 𝐻1 : 𝜇 > 41 (which is a right tailed test)

Page 5 of 19
 LECTURE 2

(ii) Check requirements: What distribution do we use for the sample test
statistic? Compute the z value of the sample test statistics 𝑥̅
Since 𝑛 ≥ 30 and we know 𝜎, we use the standard normal distribution.
Using 𝑥̅ = 47 from the sample, 𝜎 = 35, 𝜇 = 41 from 𝐻0 , 𝑎𝑛𝑑, 𝑛 = 40

𝑥̅ − 𝜇 47 − 41
𝑍= = = 1.08
𝜎/√𝑛 35⁄
√40
(iii) Find the p-value of the test statistic
Since we have a right-tailed test, the p-value is the area to the right of
𝑧 = 1.08
P-value = 𝑃(𝑍 > 1.08) = 1 − 𝑃(𝑍 ≤ 1.08)
= 1 − Φ(1.08)
= 1 − 0.85993
= 0.14007
≈ 0.1401

(iv) Conclude the test

Since the p-value of 0.1401 > 0.05 for 𝛼, we fail to reject 𝐻0
(v) Interpretation; interpret the results in the context of the problem
At the 5% level of significance, the evidence is not sufficient to reject 𝐻0 .
Based on the sample data, we do not think the average sunspot activity
during the Spanish colonial period was higher than the long-term mean.

Example 2

The Environment protection Agency has been studying Miller creek regarding ammonia
nitrogen concentration. For many years, the concentration has been 2.3mg/L. however, a
new golf course and new housing development are raising concern that the concentration

Page 6 of 19
 LECTURE 2

may have changed because of lawn fertilizer. Any change (either an increase or decrease)
in the ammonia nitrogen concentration can affect plant and animal life in and around creek
(Reference: EPA Report 832-R-93-005)
Let x be a random variable representing ammonia nitrogen concentration (in mg/L).
Based on recent studies of miller creek, we may assume that 𝑥 has a normal distribution
with 𝜎 = 0.30. Recently, a random sample of eight water tests from creek gave the
following 𝑥 values
2.1 2.5 2.2 2.8 3.0 2.2 2.4 2.9

The sample mean is 𝑥̅ = 2.51

Let us construct a statistical test to examine the claim that the concentration of ammonia
nitrogen has changed from 2.3 mg/L. Use level of significance 𝛼 = 0.01
(i) What is the null hypothesis? What is the alternative hypothesis? What is the
level of significance 𝛼?

Solution

Note that the claim to be tested is 𝜇 ≠ 2.3 𝑚𝑔/𝐿

Note that the symbolic form that must be true when the original claim is false is 𝜇 =
2.3 𝑚𝑔/𝐿
Now identify the null and alternative hypotheses. The null hypothesis (denoted by 𝐻0 ) is
a statement that the value of a population parameter (such as proportion, mean or
standard deviation) is equal to some claimed value.
The alternative hypothesis (denoted by 𝐻1 ) is a statement that the parameter has a value
that somehow differs from the null hypothesis
Consider the two symbolic expressions obtained so far, 𝜇 ≠ 2.3𝑚𝑔/𝐿 and 𝜇 = 2.3𝑚𝑔/𝐿.
The alternative hypothesis 𝐻1 is the one not containing equality. The null hypothesis 𝐻0 is
the symbolic expression that the parameter equals the fixed value being considered. The
null hypothesis and alternative hypothesis are;
𝐻0 : 𝜇 = 2.3𝑚𝑔/𝐿
𝐻𝑎 : 𝜇 ≠ 2.3 𝑚𝑔/𝐿

(ii) Is this a right-tailed, left-tailed, or two-tailed test?

Solution

To determine the type of the test, look at the symbol used in the alternative
hypothesis(𝐻1 ).
The symbol > points to the right and the test is a right-tailed

Page 7 of 19
 LECTURE 2

The symbol < point to the left and the test is left-tailed
The symbol ≠ is used for a two-tailed test.
Since the alternative hypothesis is 𝐻𝑎 : 𝜇 ≠ 2.3 𝑚𝑔/𝐿, this is a two-tailed test.

(iii) Check requirements: what sampling distribution shall we use? Note that the
value of 𝜇 is given in the null hypothesis, 𝐻0

Solution

Since the 𝑥 distribution is normal and 𝜎 is known, we use the standard normal distribution
with
𝑥̅ − 𝜇 𝑥̅ − 2.3
𝑍=𝜎 =
⁄ 𝑛 0.3⁄
√ √8

(iv) What is the value of the sample test statistic? Convert the sample mean 𝑥̅ to
a standard 𝑍 value

Solution

The sample of eight measurements has mean 𝑥̅ = 2.51.

Converting this measurement to 𝑍, we have
2.51 − 2.3
𝑍= = 1.98
0.3⁄
√8

(v) Draw a sketch showing the p-value area on the standard normal
distribution. Find the p-value

Solution

P-value = 2𝑃(𝑍 > 1.98) = 2(1 − 𝑃(𝑍 ≤ 1.98))

= 2(1 − Φ(1.98)) = 2(1 − 0.97615)
= 0.0477

Page 8 of 19
 LECTURE 2

(vi) Compare the level of significance 𝛼 and the p-value. What is your
conclusion?

Solution

Since p-value 0.0477 ≥ 0.01, we fail to reject 𝐻0

(vii) Interpret your results in the context of this problem.

Solution

The sample data are not significant at 𝛼 = 0.01 level. At this point in time, there is not
enough evidence to conclude that the ammonia nitrogen has changed in miller creek.

 In most statistical applications, the level of significance is specified to be 𝛼 =

0.05 𝑜𝑟 𝛼 = 0.01, although other values can be used.
 If 𝛼 = 0.05, then we say we are using a 5% level of significance
 This means that in 100 similar situations, 𝐻0 will be rejected 5 times on average
when it should not have been rejected.
 When we accept ( or fail to reject) the null hypothesis, we should understand that we
are not proving the null hypothesis, but, we are saying only that the sample
evidence(data) is not strong enough to justify the rejection of the null hypothesis
 The word accept sometimes has a stronger meaning in common English usage than we
are willing to give it in our application of statistics.
 Therefore, we often use the expression “ fail to reject” the null hypothesis instead of
“accept 𝑯𝟎 ”

Page 9 of 19
 LECTURE 2

 “Fail to reject” the null hypothesis simply means the evidence in favor of rejection was
not strong enough
 Often, in the case that 𝐻0 cannot be rejected , a confidence interval is used to estimate
the parameter in question
 The confidence interval gives the statistician a range of possible values for the
parameter.

Meaning of the Terms fail to reject 𝑯𝟎 and reject 𝑯𝟎

Term Meaning
Fail to Reject 𝐻0 There is not enough evidence in the data (and the test being employed) to
justify a rejection of 𝐻0 . This means that we retain 𝐻0 with the
understanding that we have not proved it to be true beyond all doubt
Reject 𝐻0 There is enough evidence in the data (and the test employed) to justify
rejection of 𝐻0. This means that we choose the alternative hypothesis
𝐻1 with the understanding that we have not proved 𝐻1 to be true beyond all
doubt.

Comment about p-value and level of significance

 The level of significance 𝛼 should be a fixed, pre specified value. Usually 𝛼 is chosen
before any samples are drawn
 The level of significance 𝛼 is the probability of a type I error. So 𝛼 is the probability of
rejecting 𝐻0 when, in fact, 𝐻0 is true.
 The p-value should not be interpreted as the probability of a type I error. The level of
significance (in theory) is set in advance before any samples are drawn.
 The P-value cannot be set in advance, since it is determined from the random sample.
 Therefore, the p-value, together with 𝛼 should be regarded as tools used to conclude
the test.
 If the p-value ≤ 𝛼 then reject 𝐻0 and say that the results are significant , otherwise,
we retain 𝐻0 and say the evidence was not sufficient to reject 𝐻0
 If p-value > 𝛼, then fail to reject 𝐻0
 If the sample evidence tells us not to reject 𝐻0, we have not proved 𝐻0 to be true
beyond all doubt. We simply do not have enough evidence to reject 𝐻0 at a specified
level of significance 𝜶 and the computed p-value
Note;
 In statistical work; the English Statistician F.Y. Edge worth (1845-1926) was one of
the first to use the term “significant” to imply that the sample data indicate a
“meaningful” difference from a previously held view.
 Thus a test is significant when 𝐻0 is rejected

Page 10 of 19
 LECTURE 2

(ii) Testing claims about a population mean (𝝁) when (𝝈) is unknown
 When the 𝜎 is unknown, we usually use the student’s 𝒕 test
Requirements
 Let x be a random variable appropriate to your application.
 Obtain a simple random sample of size 𝑛 of 𝑥 values from which you can
compute the sample mean 𝑥̅ and the sample standard deviation s. if you
can assume 𝑥 has a normal distribution or simply a mound-shaped and
symmetric distribution, then sample size 𝑛 will work.
 If you cannot assume this, use a sample size 𝑛 ≥ 30.
Testing procedure
(i) In the context of the application, state the null and alternative
hypotheses and set the level of significance 𝛼
(ii) Use 𝑥̅ , 𝑠 and 𝑛 from the sample with 𝜇 from 𝐻0 , to compute the 𝑡 value
𝑥̅ −𝜇
𝑡=𝑠 With degrees of freedom 𝑑𝑓 = 𝑛 − 1
⁄ 𝑛
√

(iii) Use the student’s t distribution and the type of test, one-tailed or two-
tailed, to find (or estimate) the p-value corresponding to test statistic.
(iv) Conclude the test. If p-value ≤ 𝛼, then reject 𝐻0 . If p-value > 𝛼 , then do
not reject 𝐻0
(v) Interpret your conclusion in the context of the application
Note
If the test statistic t for the sample statistic 𝑥̅ is negative, look up the p-value for
the corresponding positive value of 𝑡 (i.e. Look up the p-value |t|)

P-value for one-tailed test and for two-tailed tests

Page 11 of 19
 LECTURE 2

Example 1

The drug 6 –MP (6-mercaptopurine) is used to treat leukemia. The following data
represent the remission times (in weeks) for a random sample of 21 patients using
6-Mp (reference: E.A Gehan, University of Texas cancer center)
10 7 32 23 22 6 16 34 32 25 11
20 19 6 17 35 6 13 9 6 10

The sample mean 𝑥̅ = 17.1 weeks, with sample standard deviation 𝑠 = 10.0. Let 𝑥
be arandom variable representing the remission time (in weeks) for all patients.
Using 6-MP.
Assume the 𝑥 distribution is mound shaped and symmetric. A previously used drug
treatment had a mean remission time of 𝜇 = 12.5 weeks. Do the data indicate that
the mean remission time using the drug 6-MP is different (either way) from 12.5
weeks? Use 𝛼 = 0.01

Solution

(i) Establish the null and alternative hypothesis

Since we want to determine if the drug 6-MP provides a mean remission
time that is different from that provided by a previously used drug having
𝜇 = 12.5 𝑤𝑒𝑒𝑘𝑠
H0 : μ = 12.5 weeks
Ha : μ ≠ 12.5 weeks

(ii) Check requirements: what distribution do we use for the sample test
statistic 𝑥̅ compute the sample test statistic 𝑥̅ and the corresponding 𝑡
value
The 𝑥 distribution is assumed to be mound-shaped and shaped and
symmetric. Because we don’t know 𝜎, we use a student’s 𝑡 distribution.
With 𝑑𝑓 = 20. Using ̅𝑥 = 17.1 and 𝑠 = 10.0 from the sample data, 𝜇 = 12.5
from 𝐻0 , and, 𝑛 = 21,
𝑥̅ − 𝜇 17.1 − 12.5
𝑡=𝑠 = = 2.108
⁄ 𝑛 10.0⁄
√ √21
(iii) Find the p-value or the interval containing the p-value
The sample statistic is 𝑡 = 2.108 and since this is a two-tailed test with
𝑑𝑓 = 20 − 1 = 19 degrees of freedom, then, the p-value for the sample t
falls between the corresponding two-tailed areas 0.050 and 0.020
0.020 < 𝑝 𝑣𝑎𝑙𝑢𝑒 < 0.050
Page 12 of 19
 LECTURE 2

(iv) Conclude the test

As the diagram shows, the entire range is greater than 𝛼. This means the
specific p-value is greater than𝛼, so we cannot reject 𝐻0

(v) Interpretation: interpret the results in the context of the problem

At the 1% level of significance, the evidence is not sufficient to reject 𝐻0 .
Based on the sample data, we cannot say that the drug 6-MP provides a
different average remission time than the previous drug

Example 2

Archaeologists become excited when they found an anomaly in discovered

artifacts. The anomaly may (or may not) indicate a new trading region or anew
method of craftsmanship. Suppose the lengths of projectile points (arrow heads)
at a certain archaeological site have mean length 𝜇 = 2.6 𝑐𝑚 . A random sample
of 61 recently discovered projectile points in an adjacent cliff dwelling gave the
following lengths (in cm) (Reference: A. Woolsey and A. McIntyre, Mimberes
Mogollon Archaeology, university of New Mexico press).
3.1 4.1 1.8 2.1 2.2 1.3 1.7 3.0 3.7 2.3 2.6 2.2 28 3.0
3.2 3.3 2.4 2.8 2.8 2.9 2.9 2.2 2.4 2.1 3.4 3.1 1.6 3.1
3.5 2.3 3.1 2.7 2.1 2.0 4.8 1.9 3.9 2.0 5.2 2.2 2.6 1.9
4.0 3.0 3.4 4.2 2.4 3.5 3.1 3.7 3.7 2.9 2.6 3.6 3.9 3.5
1.9 4.0 4.0 4.6 1.9

Page 13 of 19
 LECTURE 2

The sample mean is 𝑥̅ = 2.92𝑐𝑚 and the sample standard deviation is 𝑠 = 0.85,
where 𝑥 is a random variable that represents the lengths (in cm) of all projectile
points found at the adjacent cliff dwelling site. Do these data indicate that the mean
length of projectile points in the adjacent cliff dwelling is longer than 2.6 cm? Use
a 1% level of significance.
(i) State H0 , H1 and α

Solution

H0 : μ = 2.6 cm vs H1 : μ > 2.6cm ; α = 0.01

(ii) Check requirements: what sampling distribution should you use for 𝑥̅ ?
what is the t-value of the sample test statistic

Solution

Because 𝑛 ≥ 30 and 𝜎 is unknown, use the student’s 𝑡 distribution with

𝑑𝑓 = 𝑛 − 1 = 61 − 1 = 60. Using 𝑥̅ = 2.92, 𝑠 = 0.85, 𝜇 = 2.6 from 𝐻0 and 𝑛 = 61
𝑥̅ − 𝜇 2.92 − 2.6
𝑡=𝑠 = = 2.940
⁄ 𝑛 0.85⁄
√ √61

(iii) Find an interval containing the p-value, do you use one-tailed or two-tail
areas? Why? Sketch a figure showing the p-value. Find an interval
containing the p-value.

Solution

This is a right-tailed test, so we use a one-tailed area

Using 𝑑𝑓 = 60, we find that the sample 𝑡 = 2.940 is between the critical values
2.660 and 3.232. The sample p-value is then between the one-tail areas 0.005
and 0.001. That is;
0.001 < 𝑝 𝑣𝑎𝑙𝑢𝑒 < 0.005

Page 14 of 19
 LECTURE 2

(iv) Do we reject of fail to reject 𝐻0 ?

Solution

Since the interval containing the p-value lies to the left of 𝛼 = 0.01, we reject 𝐻0

(v) Interpretation: interpret your results in the context of the application

Solution

At the 1% level of significance, sample evidence is sufficiently strong to reject 𝐻0

and conclude that the average projectile point length at the adjacent cliff
dwelling site is longer than 2.6cm

Testing 𝝁 using critical regions (Traditional Method)

 The most popular method of statistical testing is the p-value method.
 Another method of testing is called the critical region method or traditional
method
 For affixed, present value of the level of significance 𝛼, both methods are
logically equivalent.
 Consider the null hypothesis 𝐻0 : 𝜇 = 𝑘. We use the information from a random
sample together with the sampling distribution for 𝑥̅ and the level of
significance 𝛼, to determine whether or not we should reject the null hypothesis
 The essential question is “How much can 𝑥̅ vary from 𝜇 = 𝑘 before we suspect
that 𝐻0 : 𝜇 = 𝑘 is false and reject it?
 The answer to the question regarding the relative sizes of 𝑥̅ and 𝜇, as stated in
the null hypothesis, depends on the sampling distribution of 𝑥̅ , the alternative
hypothesis 𝐻1 and the level of significance 𝛼.
 If the sample statistic 𝑥̅ is sufficiently different from the claim about 𝜇 made in
the null hypothesis, we reject the null hypothesis.

Page 15 of 19
 LECTURE 2

 ̅
The values of 𝑥̅ for which we reject the 𝐻0 are called the critical region of the 𝒙
distribution
 Depending on the alternate hypothesis, the critical region is located on the left-
side, on the right side or both sides of the 𝑥̅ distribution.
 The figure below sows the relationship of the critical region to the alternate
hypothesis and the level of significance 𝛼

 Notice that the total area in the critical region is preset to be the level of
significance 𝜶
 Recall that the level of significance 𝛼 should be in theory be a fixed, preset
number assigned before drawing any samples
 The most commonly used levels of significance are 𝛼 = 0.05 and 𝛼 = 0.01
 Critical values are the boundaries of the critical region.
 Critical values designated as 𝑍0 for the standard normal distribution are shown
below

Page 16 of 19
 LECTURE 2

 The procedure for hypothesis testing using the critical region follows the same
procedure as the one for the p-value

Procedure for testing using the critical region approach

Requirements
 Let 𝑥 be random variable appropriate to your application.
 Obtain a simple random sample of size 𝑛 of the 𝑥 values from which you
compute the sample mean 𝑥̅ . The value of 𝜎 is already known perhaps from
the previous study. If you can assume that 𝑥 has a normal distribution then any
sample size 𝑛 will work
 If you cannot assume this use a sample size 𝑛 ≥ 30. Then 𝑥̅ follows a
distribution that is normal or approximately normal
Procedure
(i) In the context of the application, state the null and alternative hypotheses
and set the level of significance 𝛼. We use the most popular choices.

Page 17 of 19
 LECTURE 2

(ii) Use the known 𝜎, the sample size 𝑛, the value of 𝑥̅ from the sample and 𝜇
from the null hypothesis to compute the standardized sample test statistic
𝑥̅ − 𝜇
𝑍=𝜎
⁄ 𝑛
√
(iii) Show the critical region and the critical values on the graph of the sampling
distribution. The level of significance 𝛼 and the alternative hypothesis
determine the locations of the critical regions and the critical values.
(iv) Conclude the test. If the test statistic Z computed in step 2 is in the critical
region, then reject the null hypothesis. If the test statistic Z are not in the
critical region, then do not reject 𝐻0.
(v) Interpret your conclusions in the context of the application

Example

Consider the example regarding sunspots. Let x be a random variable representing the
number of sunspots observed in a -4 week period. A random sample of 40 such periods
from Spanish colonial gave the number of sunspots per period. The raw data are given
previously. The sample mean is 𝑥̅ = 47
Previous studies indicate that for this period, 𝜎 = 35. It is thought that for thousands of
years, the mean number of sunspots per 4-week period was about 𝜇 = 41. Do the data
indicate that the mean sunspot activity during the Spanish colonial period was higher than
41? Use 𝛼 = 0.05

Solution

(i) Set the null and alternate hypothesis

As from the previous example, we use
𝐻0 : 𝜇 = 41 and 𝐻𝑎 : 𝜇 > 41
(ii) Compute the Z value of the sample test statistic as in example 3, we use the
standard normal distribution with 𝑥̅ = 47, 𝜎 = 35, 𝜇 = 41 from 𝐻0 and 𝑛 =
40
𝑥̅ − 𝜇 47 − 41
𝑍= 𝜎 = = 1.08
⁄ 𝑛 35⁄
√ √40
(iii) Determine the critical region and the critical values based on 𝐻1 and 𝛼 =
0.05 . Since we have a right-tailed test, the critical region is the rightmost
5% of the standard normal distribution. According the standard normal
table, the critical value is 𝑧0 = 1.645
(iv) Conclude the test

Page 18 of 19
 LECTURE 2

We conclude that the test by showing the critical region, critical value, and
the sample statistic 𝑧 = 1.08 on the standard normal curve. For a right-
tailed test with 𝛼 = 0.05 the critical value is 𝑧0 = 1.645
The figure below shows the critical region. As we can see, the sample test
statistic does not fall in the critical region. Therefore, we fail to reject the
null hypothesis.

(v) Interpretation: interpret the results in the context of the application

At the 5% level of significance, the sample evidence is insufficient to justify
rejecting 𝐻0 . It seems that the average sunspot activity during the Spanish
colonial period was the same as the historical average

(vi) How do results of the critical region method compare to the results of the p-
value method for a 5% level of significance?
The results as expected, are the same. In both cases we fail to reject to reject
𝐻0

Page 19 of 19