STA 204

Statistics for Agricultural

& Biological Sciences II
Dr. M. K. Garba Caution
This learning material is NOT
Department of Statistics, FOR SALE by anybody, not even
University of Ilorin, Nigeria by the Class Representative.
08035568461 Appropriate sanctions await the
violators.

https://groups.google.com/a/unilorin.edu.ng/d/forum/sta204_2018
 Common Significance Levels and their
Corresponding Z-Values
The following are the four commonly used
significance levels in hypotheses testing and their
associated Z-values.
α Zα α/2 Zα/2
One-tailed test

Two-tailed test
0.10 1.282 0.05 1.645
0.05 1.645 0.025 1.96
0.02 2.054 0.01 2.33
0.01 2.33 0.005 2.575
2
This learning material is NOT FOR SALE
 Test for a Population Proportion
In certain situation, particularly with count data,
we may be interested in proportions rather than
means of population. In such cases, we carry out
test of hypothesis that the sample proportion
could have come from a population with a given
proportion.
We may wish to test the hypothesis that the
population proportion P is a given value P0.

This learning material is NOT FOR SALE 3

Such hypotheses are sated as

𝑯𝟎 ∶ 𝑷 = 𝑷𝟎 𝒗𝒔 𝑯𝟏 ∶ 𝑷 < 𝑷𝟎

𝑯𝟎 ∶ 𝑷 = 𝑷𝟎 𝒗𝒔 𝑯𝟏 ∶ 𝑷 > 𝑷𝟎

𝑯𝟎 ∶ 𝑷 = 𝑷𝟎 𝒗𝒔 𝑯𝟏 ∶ 𝑷 ≠ 𝑷𝟎

This learning material is NOT FOR SALE 4

Test Statistic for a Population Proportion
To estimate a population proportion (P), we assume
that a sample of size n is drawn from a population of
interest and the number of subjects having the
characteristics of interest is observed (say x). Thus x
subjects have the characteristics of interest. An
estimate of the population proportion P is therefore
𝒙
𝒑= such that
𝒏
𝒑𝒒 𝒑(𝟏−𝒑)
E 𝒑 = 𝑷 and 𝑽 𝒑 = 𝒐𝒓
𝒏 𝒏
5
This learning material is NOT FOR SALE
The test statistic is given as

𝒙 − 𝒏 𝒑𝟎
𝒁=
𝒏 𝒑𝟎 𝟏 − 𝒑𝟎
or equivalently as
𝒑 − 𝒑𝟎
𝒁=
𝒑𝟎 𝟏 − 𝒑𝟎
𝒏
𝒙
𝒔𝒊𝒏𝒄𝒆 𝒑 = then 𝒙 = 𝒏𝒑
𝒏
6
This learning material is NOT FOR SALE
 Decision Rules
1. For one-tailed tests, reject the null hypothesis
in favour of its alternative if Zcalculated is
greater or equal to Ztabulated (i.e. Zcal ≥ Zα),
else do not reject the null hypothesis.

2. For two-tailed tests, reject the null hypothesis

in favour of its alternative if Zcalculated is
greater or equal to Ztabulated (i.e. Zcal ≥ Zα/2),
else do not reject the null hypothesis.

7
This learning material is NOT FOR SALE
 Example 17
A recent survey carried out at UITH indicates
that out of 3,232 births recorded this year,
1705 were girls and the rest were boys .
Do these figures confirm the hypothesis that
the sex ratio is 50:50? Perform your test at
5% level of significance.

This learning material is NOT FOR SALE 8

 Example 18
The Agent for Arik airline claimed that 20
percent of the passengers travel in first class, but
the General Manager felt there is possibility that
this percentage could be more or less. Then he
randomly sampled 400 passengers of which 70
passengers holding first class tickets. Can the
null hypothesis as stated by the Agent be rejected
at 10% level of significance?

This learning material is NOT FOR SALE 9

 Exercise
In a certain cross of two varieties of peas, genetic
theory made the researcher to expect one-half of
the seeds produced to be wrinkled and the
remaining one half to be smooth. In order to test
this hypothesis, a final-year student of Plant
Biology conducted the experiment with 40 seeds
of the cross varieties and observed that 30 are
wrinkled and 10 are smooth. Assuming α = 0.05,
is the genetic theory right?

This learning material is NOT FOR SALE 10

HYPOTHESIS TESTS
CONCERNING TWO
POPULATIONS

11
This learning material is NOT FOR SALE
 Case I: Independent Samples
If samples of sizes n1 and n2 are drawn from two
independent populations whose variances σ1 and σ2
are known, we may want to test whether significant
differences exist between the means of the two
populations. To do this, we carry out a statistical test
of which the hypotheses are as follows
H0: μ1 = μ2 or H0: μ1 – μ2 = 0
Versus
H1: μ1 ≠ μ2 or H1: μ1 – μ2 ≠ 0
12
This learning material is NOT FOR SALE
While the hypotheses for one-tailed tests are
H0: μ1 = μ2 or H0: μ1 – μ2 = 0
Lower-tailed test

Versus
H1: μ1 < μ2 or H1: μ1 – μ2 < 0
Upper-tailed test

H0: μ1 = μ2 or H0: μ1 – μ2 = 0
Versus
H1: μ1 > μ2 or H1: μ1 – μ2 > 0
13
This learning material is NOT FOR SALE
 Test Statistics
1. Population variances are known and the samples
happen to be large.
Distributed as Zα or Zα/2
𝒙𝟏 − 𝒙𝟐 − 𝝁𝟏 − 𝝁𝟐 depending on whether the
𝒁= test is one-tailed or 2-tailed
𝝈𝟐
𝟏 𝝈𝟐
+ 𝟐
𝒏𝟏 𝒏𝟐

2. Population variances are not known and the

samples happen to be small.
𝒙𝟏 − 𝒙𝟐 − 𝝁𝟏 − 𝝁𝟐
𝒕=
𝟏 𝟏 Distributed with n1+n2 -2
𝑺𝟐
𝒑 −
𝒏𝟏 𝒏𝟐 degrees of freedom as tα or tα/2
depending on whether the test is
one-tailed or 2-tailed
𝟐 𝟐
𝒏𝟏 − 𝟏 𝑺𝟏 + 𝒏𝟐 − 𝟏 𝑺𝟐
𝒘𝒉𝒆𝒓𝒆 𝑺𝟐𝒑 =
𝒏𝟏 + 𝒏𝟐 − 𝟐 14
This learning material is NOT FOR SALE
 Example 19
In a random samples of 12 observations from each
of two normal populations, we found the following
statistics

𝒙𝟏 = 𝟕𝟒, 𝒙𝟐 = 𝟕𝟏, 𝒔𝟏 = 𝟏𝟖 𝒂𝒏𝒅 𝒔𝟐 = 𝟏𝟔

Test at α = 0.05 to determine whether it can be
inferred that the population means differ.

This learning material is NOT FOR SALE 15

 Example 20
Random sampling from two normal populations produced
the following datasets. Test whether the two datasets differ
at 5% significance level.
Sn 1 2 3 4 5 6 7 8 9 10 11 12
Set A 6.6 5.7 5.6 6.4 4.8 6.3 6.0 7.3 5.9 6.4 9.0 7.0

Set B 6.3 7.0 8.2 6.1 6.7 5.9 6.1 7.3 6.6 4.5 4.9 7.9

13 14 15 16 17 18 19 20 21 22 23 24 25

5.2 6.3 8.1 8.6 5.0 7.1 8.1 6.1 6.3 4.5 6.1 7.3 6.1

3.3 5.7 7.9 4.5 5.5 5.7 5.0 5.6 6.4 6.7 6.7 7.3 6.5
This learning material is NOT FOR SALE 16
 Exercise
Given the following summaries for two
independent populations from which 12 samples
were taking from each population, test at 5%
significance level whether the two population
means are different.

𝒙𝟏 = 𝟐𝟎. 𝟏𝟕, 𝒙𝟐 = 𝟏𝟔. 𝟐𝟑, 𝒔𝟏 = 𝟒. 𝟑 𝒂𝒏𝒅 𝒔𝟐 = 𝟑. 𝟖

This learning material is NOT FOR SALE 17

 Case II: Paired Samples
This test of hypothesis is otherwise known as the
paired t-test. It is used to compare two population
means in which case we have pairs of observations
from each sample point.
Paired samples may arise in a number of situations.
The following are some of those situations:
 A particular treatment/drug may be applied to a
set of experimental units and the readings or
observations on each unit before and after the
experiment are recorded 18
This learning material is NOT FOR SALE
 Many sets of twins may be made available for an
experiment and two types of treatments be
administered on each set of the twins with an
individual receiving a treatment, then observations
are recorded.
 A number of plots of land may be acquired in
different locations and each plot be divided into
halves (2 equal-size sub-plots) and one of which is
treated with different brands of fertilizer. Then the
yields or responses are recorded.
The observations from these scenarios are referred to as
paired samples, matched pairs or related samples.
19
This learning material is NOT FOR SALE
The hypotheses to may be tested are
𝑯𝟎 : 𝝁𝒅 = 𝟎 𝒗𝒔 𝑯𝟏 : 𝝁𝒅 < 𝟎

𝑯𝟎 : 𝝁𝒅 = 𝟎 𝒗𝒔 𝑯𝟏 : 𝝁𝒅 > 𝟎

𝑯𝟎 : 𝝁𝒅 = 𝟎 𝒗𝒔 𝑯𝟏 : 𝝁𝒅 ≠ 𝟎

for one-tailed and two-tailed tests

respectively.
This learning material is NOT FOR SALE 20
The test statistic is
𝒙𝒅 − 𝝁𝒅
𝒕=
𝒔𝒅 𝒏𝒅
which is distributed as Zα (if it is one-tailed) or
Zα/2 (if it is two tailed) when the sample size is
large. But when the sample size is small, the
test statistic is distributed as 𝒕𝒏𝒅 − 𝟏, α or
𝜶 for one-tailed or two-tailed test.
𝒕𝒏𝒅 − 𝟏,
𝟐
This learning material is NOT FOR SALE 21
 Example 21
To determine the efficiency of a new method of
teaching a particular topic, a random sample of 12
students were given a diagnostic test before teaching
the topic and another test after the teaching. The cores
of the students are given below. Can it be inferred that
the new method is efficient? Take α =5%

Before 68 44 30 58 35 33 52 69 23 69 48 30

After 59 42 20 62 25 30 56 62 25 75 40 26

This learning material is NOT FOR SALE 22

 Exercise
Memory capacity of 9 students was tested before
and after training. Verify at 5% level of significance
whether the training was effective using the
following scores

Before 10 15 9 3 7 12 16 17 4

After 12 17 8 5 6 11 18 20 3

This learning material is NOT FOR SALE 23

#1
PRACTICE QUESTIONS
An investigator randomly selected 36 nerve cells from a certain region of the brain of male
guinea pigs. The counted number of dendritic branch segments emanating from each selected
cell are as follows:

38 42 25 35 35 33 48 53 17
24 26 26 47 28 24 35 38 26
38 29 49 26 41 26 35 38 44
25 45 28 31 46 32 39 59 53

The mean 𝒙 for these counts is 35.67 and the sample standard deviation (s) is 9.99.
(a) Construct a 95 % confidence interval for the population mean (μ) the number of dendritic
segment counts that can emanate from the body of a male pig nerve cell.
(b) Test the hypothesis at α = 0.05 that: H0: μ = 30 vs H1: μ > 30
(c) Interpret your results in (a) and (b) to a layman

This learning material is NOT FOR SALE 24

 #2
Among patients with lung cancer, usually 90 % or more die within 3 years. As a result of new
forms of treatment, it is felt that this rate has been reduced. In a recent study of 150 patients
diagnosed with lung cancer, 128 died within 3 years.
(a) Calculate a point estimate of p, the true proportion of lung cancer patients who died within 3
years.
(b) Set up the null and alternative hypotheses needed to support the above contention.
(c) Can H0 be rejected at the α = 0.10 level of significance?
(d) Can H0 be rejected at the α = 0.05 level of significance?
(e) Do you think that there is sufficient evidence to claim that the new methods of treatment are
more effective than the old? Explain.

This learning material is NOT FOR SALE 25