Approaches to Hypothesis Testing

There are three approaches to hypothesis testing namely Classical Approach, p vale approach and
the confidence interval approach

The Classical Approach

The Classical Approach to hypothesis testing is to compare a test statistic and a critical value. It
is best used for distributions which give areas and require you to look up the critical value (like
the Student's t distribution) rather than distributions which have you look up a test statistic to find
an area (like the normal distribution).
The Classical Approach also has three different decision rules, depending on whether it is a left
tail, right tail, or two tail test.
One problem with the Classical Approach is that if a different level of significance is desired, a
different critical value must be read from the table.

P-Value Approach
The P-Value, short for Probability Value, Approach to hypothesis testing form a different
manner. Instead of comparing z-scores or t-scores as in the classical approach, you're comparing
probabilities, or areas.
The level of significance (alpha) is the area in the critical region. That is, the area in the tails to
the right or left of the critical values.
The p-value is the area to the right or left of the test statistic. If it is a two tail test, then look up
the probability in one tail and double it.
If the test statistic is in the critical region, then the p-value will be less than the level of
significance. It does not matter whether it is a left tail, right tail, or two tail test. This rule always
holds.
Reject the null hypothesis if the p-value is less than the level of significance.
You will fail to reject the null hypothesis if the p-value is greater than or equal to the level of
significance.
The p-value approach is best suited for the normal distribution when doing calculations by hand.
However, many statistical packages will give the p-value but not the critical value. This is
because it is easier for a computer or calculator to find the probability than it is to find the critical
value.
Another benefit of the p-value is that the statistician immediately knows at what level the testing
becomes significant. That is, a p-value of 0.06 would be rejected at an 0.10 level of significance,
but it would fail to reject at an 0.05 level of significance. Warning: Do not decide on the level of
significance after calculating the test statistic and finding the p-value.
Here are a couple of statements to help you keep the level of significance the probability value
straight.
The Level of Significance is pre-determined before taking the sample. It does not depend on the
sample at all. It is the area in the critical region; that is the area beyond the critical values. It is
the probability at which we consider something unusual.
The Probability-Value can only be found after taking the sample. It depends on the sample. It is
the area beyond the test statistic. It is the probability of getting the results we obtained if the null

1
 hypothesis is true.

Confidence Intervals as Hypothesis Tests

Using the confidence interval to perform a hypothesis test only works with a two-tailed test.
a) If the hypothesized value of the parameter lies within the confidence interval with a 1-a level
of confidence, then the decision at a level of significance is to fail to reject the null
hypothesis.
b) If the hypothesized value of the parameter lies outside the confidence interval with a 1-a
level of confidence, then the decision at a level of significance is to reject the null
hypothesis.
However, it has a couple of problems.
 It only works with two-tail hypothesis tests.
 It requires that you compute the confidence interval first. This involves taking a z-score or t-
score and converting it into an x-score, which is more difficult than standardizing an x-
score.
Testing a Single Mean
The value for all population parameters in the test statistics come from the null hypothesis. This
is true not only for means, but all of the testing we're going to be doing.
The following hypotheses are to be tested:
H 0 : μ=μ0 vs H 1 : μ≠μ 0 0r H 0 : μ> μ 0 0r
H 0 : μ< μ 0 Where μ0 is some hypothesised value.
The statistic and the critical values depends on whether σ, is known or unknown.

Hypoth

σ esis σ Unknown
Known
Known Unknown
(Z test) (t test)
Tests
The test
for 
statistic is:
a) Population Standard Deviation Known
If the population standard deviation σ, is known, then the population mean has a normal
distribution, and you will be using the z-score formula for sample means. The test statistic is the
x−μ
z=
standard formula you've seen before. σ / √n
The critical value is obtained from the normal table.

Example 1
Test at 5% level the claim that the true mean # of TV sets in US homes is equal to 3. Suppose the
sample results are n = 100, x=2.84 (σ = 0.8 is assumed known)
Solution

2
 State the appropriate null and alternative hypotheses
- H0: μ = 3 H1: μ ≠ 3 (This is a two-tail test)
Determine the appropriate technique
- σ is assumed known so this is a Z test.
Determine the critical values
- For a = 0.05 the critical Z values are ±1.96
Compute the test statistic So the test statistic is:
X − μ 2 .84 − 3
ZSTAT = σ/ √ n
= 0. 8/ √ 100 =−2 .0

= =
0.05/
Reject Do not
0.05/
Reject
2-ZH0 = 0 H0+Z
reject =2
H0
α/2 α/2
-
-1.96 +1.96
Since ZSTAT 2.0= -2.0 < -1.96,
reject the null hypothesis and
conclude there is sufficient
evidence that the mean number of
TVs in US homes is not equal to 3

Exercise
1. A simple random sample of 10 people from a certain population has a mean age of 27. Can we
conclude that the mean age of the population is less than 30? The variance is known to be 20.
Let = .05.
2. Bon Air Elementary School has 300 students. The principal of the school thinks that the average
IQ of students at Bon Air is at least 110. To prove her point, she administers an IQ test to 20
randomly selected students. Among the sampled students, the average IQ is 108. Assuming
variance is known to be 100, should the principal accept or reject her original hypothesis? at 5%
level of significance
3. The manager at the Omni Fitness Club in Muskegon, Michigan, believes that the recent
remodeling project has greatly improved the club’s appeal for members and that they now stay
longer at the club per visit than before the remodeling. Studies show that the previous mean time
per visit was 36 minutes, with a standard deviation equal to 11 minutes. A simple random sample
of n=200 visits is selected, and the current sample mean is 36.8 minutes. To test the manager’s

3
 claim, and partially justify the remodeling project, using ∝=0.05 level, the following steps can
be used:
4. The Wilson Glass Company has a contract to supply plate glass for home and commercial
windows. The contract specifies that the mean thickness of the glass must be 0.375 inches. The
standard deviation, σ, is known to be 0.05 inch. Before sending the first shipment, Wilson
managers wish to test whether they are meeting the requirements by selecting a random sample
of n=100 thickness measurements.
5. Central bank believes that if consumer confidence is too high, the economy risks over heating.
Low confidence is a warning that rcession might be on the way. In either case, the bank may
choose to intervene by altering interest rates. The ideal value for the bank's chosen measure is 50.
We may assume the measure is normally distributed with standard deviation 10. The bank takes a
survey of 25 people. Which returned a sample mean of 54 for the index. What would you advice
the bank to do? Use = .05.
6. A manager will switch to a new technology if the production process exceeds 80 units per hour.
The manager asks the company statistician to test the null hypothesis: H0: μ = 80 against the
alternative hypothesis: H1: μ >80 If there is strong evidence to reject the null hypothesis then the
new technology will be adopted. Past experience has shown that the standard deviation is 8. A
data set with n = 25 for the new technology has a sample mean of:83 Does this justify adoption
of the new technology?

b) Population Standard Deviation Unknown

If the population standard deviation σ, is unknown, then the population mean has a student's t
distribution, and you will be using the t-score formula for sample means. The test statistic is very
similar to that for the z-score, except that sigma has been replaced by s and z has been replaced
x−μ
t=
by t. ie s/√n
The critical value is obtained from the t-table. The degrees of freedom is n-1.

Example 1
A fertilizer mixing machine is set to give 12 kg of nitrate for every 100kg bag of fertilizer. Ten
100kg bags are examined. The percentages of nitrate are as follows: 11, 14, 13, 12, 13, 12, 13,
14, 11, 12. Is there reason to believe that the machine is defective at 5% level of significance?
Solution
Hypothesis H0: μ = 12 H1: μ ≠ 12 (This is a two-tail test)

σ is known so this is a t test. use the unbiased estimator ie

Critical Region based on a = 0.05 and 9 degrees freedom
s=
√ ∑ ( x−x )2
n−1

t 9 , 0.025=2.262 ie reject H : μ = 12 if |t c|≥2.262

0

From calculator x=12.5 and s=1. 0801

x −μ 12 .5−12
t c= = =1 . 4639
Test statistic s/ √ n 1. 0801/ √ 10
|t |=1.2639<2.262 , we fail to reject H and conclude that the machine is not
Decision since c 0

defective.

4
 Example 2
The following figures give the end of year profits of ten randomly selected Chemists in Nairobi
county.

On the basis of this data, test whether the average profit is greater than 30M KSH at 1% level of
significance
Solution
Hypothesis H0: μ = 30 H1: μ > 30 (This is a 1-tail test)

σ is known so this is a t test. use the unbiased estimator ie

Critical Region based on a = 0.01 and 9 degrees freedom
s=
√ ∑ ( x−x )2
n−1

t 9 , 0.01=2.82 ie reject H : μ = 12 if |t c|≥2.82

0

From calculator x=29.415 and s=3. 6601

x −μ 29 . 415−30
t c= = =−0 . 51
Test statistic s/ √ n 3 .6601/ √ 10
Decision since c
|t |=0.51<2.82 , we don’t reject H and conclude that the average profit is not
0

greater than 30M KSH.

Exercise
1. Identify the critical t value for each of the following tests:
a.A two-tailed test with a=0.05 and 11 degrees of freedom
b.A one-tailed test with a=0.01 and n=17
2. Consider a sample with n=20 x=8 .0 and s=2 Do the following hypothesis tests.
a) . H0: μ =8.7 H1: μ > 8.7 at a=0.01 b) H0: μ =8.7 H1: μ ≠ 8.7 at a=0.05
3. It is widely believed that the average body temperature for healthy adults is 98.6 degrees
Fahrenheit. A study was conducted a few years go to examine this belief. The body temperatures
of n = 130 healthy adults were measured (half male and half female). The average temperature
from the sample was found to be x=98.249 with a standard deviation s = 0.7332. Do these
statistics contradict the belief that the average body temperature is 98.6? test at 1% level of
significance
4. A study is to be done to determine if the cognitive ability of children living near a lead smelter is
negatively impacted by increased exposure to lead. Suppose the average IQ for children in the
United States is 100. From a pilot study, the mean and standard deviation were estimated to be
x=89 and s = 14.4 respectively. Test at 5% level whether there is a negative impact.
5. The average cost of a hotel room in New York is said to be $168 per night. To determine if this is
true, a random sample of 25 hotels is taken and resulted in x=$172.5 and s=$ 15 . 40 . Test the
appropriate hypotheses at a = 0.05.

a) An earlier study reported that the mean shoot length is 15cm.

b) Test whether the experimental data confirms the old view at 5% level of significance.

5
6. A sample of eleven plants gave the following shoot lengthsA simple random sample of 14 people
from a certain population gives body mass indices as shown in Table 7.2.1. Can we conclude
that the BMI is not 35? Let = .05.
subject 1 2 3 4 5 6 7 8 9 10 11 12 13 14
BM 23 25 21 37 39 21 23 24 32 57 23 26 31 45
7. A company selling licenses for a franchise operation claims that, in the first year, the yield on an
initial investment is 10%. Test the company’s claim. How should a hypothesis test be stated? If
there is strong evidence that the mean return on the investment is below 10% this will give a
cautionary warning to a potential investor. Therefore, test the null hypothesis: H0 : μ = 10 against
the alternative hypothesis: H1 : μ < 10 From a sample of n = 10 observations, the sample statistics
are: x=8 . 82 and s = 2.40
8. We know the distance that an athlete can jump is normally distributed but we do not know the
standard deviation. We record 15 jumps: 7.48 7.34 7.97 5.88 7.48 7.67 7.49 7.48 8.51 5.79
7.13 6.80 6.19 6.95 5.93 Test whether these values are consistent with a mean jump length of
7m. Do you have any reservations about this test?
9. The manufacturing process should give a weight of 20 ounces. Does the data show evidence that
the process is operating correctly? Test the null hypothesis: H0 : μ = 20 the process is operating
correctly against the alternative: H1 : μ ¹ 20 the process is not operating correctly From the data
set, the sample statistics are: n = 9, x= 20.356 (ounces) and s = 0.6126

Inferences from Two Samples

Definitions
Dependent Samples: Samples in which the subjects are paired or matched in some way. Dependent
samples must have the same sample size, but it is possible to have the same sample size without being
dependent.
Independent Samples: Samples which are independent when they are not related. Independent samples
may or may not have the same sample size.
Pooled Estimate of the Variance: A weighted average of the two sample variances when the
variances are equal. The variances are "close enough" to be considered equal, but not exactly the
same, so this pooled estimate brings the two together to find the average variance.

Two-Sample Hypothesis Testing for the Mean

Two-sample hypothesis testing is statistical analysis designed to test if there is a difference between
two means from two different populations. For example, a two-sample hypothesis could be used to
test if there is a difference in the mean salary between male and female doctors in the New York City
area. A two-sample hypothesis test could also be used to test if the mean number of defective parts
produced using assembly line A is greater than the mean number of defective parts produced using
assembly line B. Similar to one-sample hypothesis tests, a one-tailed or two-tailed test of the null
hypothesis can be performed in two-sample hypothesis testing as well. The two-sample hypothesis
test of no difference between the mean salaries of male and female doctors in the New York City area
is an example of a two-tailed test. The test of whether or not the mean number of defective parts
produced on assembly line A is greater than the mean number of defective parts produced on
assembly line B is an example of a one-tailed test. The following section provides step-by-step
instructions for performing a two-sample test of a hypothesis in Excel.

The hypothesis to be tested may be any of the following

6
 H0: μ1 H0: μ1 H0: μ1
=  
Regardless of which hypotheses used, μ1 = μ2 is always assumed to be true.

The sampling distribution involves two means, so it is called the sampling distribution of the
difference between means. The value that we are interested is the difference between the means, that
is:
x 1−x 2

The mean and standard deviation of the sampling distribution of the difference between means are
μ −μ S
given by: 1 2 and x1− x2 respectively. The latter is called the standard error of the difference
between means. Since the sample standard deviation is again used to estimate the population value,

the sampling distribution of the difference between means will also be distributed as t. So if
H 0 is
x1 −x 2
t=
S x −x
true the formula becomes: 1 2

All we need to do now is determine the formula for the standard error. However, this formula differs
depending whether we are dealing with independent or dependent groups. With the independent
groups design, the subjects in each of the two groups are different and unrelated in any way. For a
dependent groups design the most common type is called a within subjects or repeated measures
design, because the same subjects (thus actually only one group) are tested twice.
There are two possible cases when testing two population means, the dependent case and the
independent case.

a). Independent Groups

The computational formulas (which will also handle unequal sample sizes) are given by
x 1 −x 2
t=

√ ( 1 1
) √ ( )
2 2
S x − x = S2p + S 2p
1 1
+ 2 (n 1−1 )S1 +(n 2−1 )S2
S p=
1 2 n1 n2 n1 n2 n1 +n 2−2
and where is the pooled
estimate of the population variance.
Since two variances are used in estimating the standard error of the difference between means, the
degrees of freedom will equal the sum of the degrees of freedom for each of the variance estimates,
that is: df =(n 1−1 )+(n2 −1)=n1 + n2−2
Example 1
A random sample of 17 police officers in Brownsville has a mean annual income of $35,800 and a
standard deviation of $7,800. In Greensville, a random sample of 18 police officers has a mean
annual income of $35,100 and a standard deviation of $7,375. Test the claim at a = 0.01 that the
mean annual incomes in the two cities are not the same. Assume the population variances are equal.

Solution Hypothesis H0: m1 = m2 vs

7
  =  =
H a:  1   2
d.f. = n1 + n20.005-3 -2 -1 0 1 2 30.005
t
(Claim) –t0 = –2.576t0 = 2.576
–2
= 17 +
18 – 2 =
33
The pooled estimate of the population variance

S p=
√ (n1 −1) S21 +(n2 −1) S22
n 1 +n2 −2
=
√(17−1)78002 +(18−1 )7375 2
17+ 18−2
x 1 −x 2
=7584 . 0355

35800−35100
t= = =0 ,273

√( ) √( )
1 1
1 1 7584 . 0355 +
Sp + 17 18
n1 n2
Therefore The test statistic
H
Decision: don’t reject 0 and conclude that there is no sufficient evidence to support the claim the
mean annual income differ.
Example 2
Big Foods Grocery has two grocery stores located in Johnston City. One store is located on First
Street and the other on Main Street and each is run by a different manager. Each manager claims that
her store's layout maximizes the amounts customers will purchase on impulse. Both managers
surveyed a sample of their customers and asked them how much more they spent than they had
planned to, in other words, how much did they spend on impulse? The following table shows the
sample data collected from the two stores.
First Street Main Street
15.78 15.19
17.73 18.22
10.61 15.38
15.79 15.96
14.22 21.92
13.82 12.87
13.45 12.47
12.86 13.96
10.82 13.79
12.85 13.74
18.4
18.57
17.79
10.83

8
Upper-level management at Big Foods Grocery wants to know if there is a difference in the mean
amounts purchased on impulse at the two stores. Test at 5% whether there is a difference in the mean
amounts purchased on impulse at the two stores.
Solution

Hypothesis
H 0 : μ1 =μ2 vs H 1 : μ1 ≠μ2 where μ is the mean amount spent on impulse

The critical values here are based on 22 (ie 10+14-2) degrees of freedom. Ie reject 0 if |t|≥2.074
H
x 1 −x 2
t=

The test statistic.

Exercise
√ S 2p
( 1 1
+
n1 n2 )
1. In a study on Serum uric acid levels of individuals with Down's syndrome and normal individuals,
2
the following data was obtained n1 =12 , x 1 =4 . 5 and S1 =1 and
n2 =15 , x 2 =3 . 4 and S 22 =2. 25 Is there a difference between the means of individuals with
Down's syndrome and normal individuals?
2. We wish to know if we may conclude, at the 95% confidence level, that smokers, in general, have
greater lung damage than do non-smokers.
Lung destructive index
n x S
smokers 16 17.5 4.4711
Non smokers 9 12.4 4.8492
3. These data were obtained in a study comparing persons with disabilities with persons without
disabilities. A scale known as the Barriers to Health Promotion Activities for Disabled Persons
(BHADP) Scale gave the data. We wish to know if we may conclude, at the 99% confidence
level, that persons with disabilities score higher than persons without disabilities.
n x S
Disabled 132 31.83 7.93
Non Disabled 137 25.07 4.80
4. The values below are the average July temperatures (C) in central England for 1987 to1996 and
from 1997 to 2006 (from HadCETdata set).
1987 to1996 data: 15.9 14.7 18.2 16.9 17.3 16.2 15.2 18.0 18.6 16.5
1997 to 2006 data: 16.7 15.5 17.7 15.5 17.2 16.0 17.6 15.8 16.9 19.7
What conclusions do you draw from these data?
5. A stock market data base contains daily closing prices for the company Johnson & Johnson for
the year 1999. For the first six months (January to June) observations are recorded for 124 trading
days. For the period July to December observations are available for 128 trading days. An
exercise of interest is to find a 95% confidence interval estimate for the difference between the
population mean closing price in the two sample periods. Stock market prices adjust rapidly to the
arrival of new information. Therefore, it is reasonable to consider that the two samples are
independent. Summary statistics for the closing prices for the two sample periods are:
n x s2
Jan-June 124 $89.96 32.54
July-Dec 128 $97.98 21.45
6. Within a school district, students were randomly assigned to one of two Math teachers - Mrs.
Smith and Mrs. Jones. After the assignment, Mrs. Smith had 30 students, and Mrs. Jones had 25

9
 students. At the end of the year, each class took the same standardized test. Mrs. Smith's students
had an average test score of 78, with a standard deviation of 10; and Mrs. Jones' students had an
average test score of 85, with a standard deviation of 15. Test the hypothesis that Mrs. Smith and
Mrs. Jones are equally effective teachers. Use a 0.05 level of significance. (Assume that student
performance is approximately normal.)
7. The Acme Company has developed a new battery. The engineer in charge claims that the new
battery will operate continuously for at least 7 minutes longer than the old battery. To test the
claim, the company selects a simple random sample of 100 new batteries and 100 old batteries.
The old batteries run continuously for 190 minutes with a standard deviation of 20 minutes; the
new batteries, 200 minutes with a standard deviation of 40 minutes. Test the engineer's claim that
the new batteries run at least 7 minutes longer than the old. Use a 0.02 level of significance.
(Assume that there are no outliers in either sample.)
8. From hospital records, we obtain the following values for these components:
Treatment Control
Average Weight 3100 g 2750 g
SD 420 425
n 75 75
With these pieces of information, test at 5% level whether the average weight of the treated
sample is higher than for the control,
9. Suppose you are a researcher interested in the factors influencing paper grading by professors.
You have a hunch (and/or previous research) might lead you to predict that papers that are typed
are rated higher than papers that are handwritten. Research to date though, has only been
correlational and thus little can be said in terms of a cause and effect relationship.
10. Suppose you have 10 freshman students currently taking English as well as an introductory
psychology course each write one paper. They should each provide two copies of their paper (one
typed and one handwritten). Next, we enlist the aid of 20 English instructors. We randomly assign
10 instructors to each of two groups. Each instructor in one group (the control group) will grade
each of the 10 papers that are hand written, while the second group (the experimental group) will
grade the same papers that are typed. Based on the following results, does typing a paper
influence the grade it receives?
Written (1) W2 Typed (2) T2
Sum 738 60,646 856 73,356
N 9 10
Mean 82.0 85.6

b) Dependent Groups
The most common type of Dependent Groups Design is also called a Within Subjects or Repeated
Measures Design, because the same subjects (thus, actually only one group) are tested twice. There is
another situation, though, in which this analysis is sometimes used. It is called the Matched Groups
Design. In this case, there are two groups, but they are matched on some variable that is highly and
positively correlated with the DV.
The idea with the dependent case is to create a new variable, D, which is the difference between the
paired values. You will then be testing the mean of this new variable.

10
Steps in paired sample Hypothesis Testing
a) State the claim mathematically ie State H0 and Ha.
b) Specify the level of significance and Identify the degrees of freedom.
c) Determine the critical value(s) and the rejection region(s). d.f. = n – 1

d=
∑d S2d =
∑ d 2−n d 2
d) Calculate d and
Sd where n and n−1
d−μd
t=
e) Find the standardized test statistic.
S d /√n
f) Make a decision to reject or fail to reject the null hypothesis.
g) Interpret the decision in the context of the original claim.

Example:
A reading center claims that students will perform better on a standardized reading test after going
through the reading course offered by their center. The table shows the reading scores of 6 students
before and after the course. At a = 0.05, is there enough evidence to conclude that the students’
scores after the course are better than the scores before the course?
Student 1 2 3 4 5 6
Score (before) 85 96 70 76 81 78
Score (after) 88 85 89 86 92 89
Solution

Hypothesis
H 0 : : μ1 =μ2 vs H 0 : : μ1 > μ2 a = 0.05 and d f=6-1=5
t
Critical value 5 , 0 .05
=2.015
Test statistics
d = (score before) – (score after)
Student 1 2 3 4 5 6
Score (before) 85 96 70 76 81 78
Score (after) 88 85 89 86 92 89 total
d -3 11 -19 -10 -11 -11 -43
7 .167−0
t c= =−1. 714
From the calculatord=−7.167 and
Sd =10. 245 implying 10 . 245/ √ 6
t =1 . 714<2 . 015 we fail to reject H 0 : and conclude that there is no sufficient
Decision: Since c
evidence at the 5% level to support the claim that the students’ scores after the course are better than
the scores before the course.

Exercise
1. Table gives B (before) and A (after) treatment data for obese female patients in a weight-loss
program.

Calculate d= A-B for each pair of data. Is the treatment effective in causing weight reduction in
these people. Test at 95% confidence level.

11
2. Suppose you are interested in reactions times to different coloured lights (especially green and red).
If a random sample of 10 subjects was tested and gave the information below Test at 5% level
whether there is a difference in reaction time for X and Y.
subject 1 2 3 4 5 6 7 8 9 10
1
X(Red) 8 16 23 30 32 30 31 25 27 21
2
Y(Green) 2 20 29 35 27 29 33 29 31 24
3 After implementing a series of °awed policies, a senior politician's approval rating is at an all time
low. He commissions a market research company to carry out a survey of 120 people to determine
their attitude toward him. He then goes on a charm offensive (being nice to old people, conspicuously
enjoying the national sport, etc.). The survey is then repeated on the same group of 120 people. We
2
calculate sample mean and variance for the differences. d=1.24 and Sd =110 . What conclusions
do you draw from these results?

12