Week No.

Topic: Inferential Statistics: Simple Test of Hypothesis - The Z-test and

the T-test Statistical Tools.

What is a hypothesis?

A hypothesis is a statement or tentative theory which aims to explain facts

about the real world. Most hypotheses have their origin in a question to some
practical problem. In search for an answer “educated guesses” and pertinent
evidences are brought out which later on are turned into propositions or hypotheses.
These hypotheses are then subjected for testing. If they are found to be
statistically true, they are accepted. If they are found to be false, they are rejected.
The hypothesis that we hope to accept or reject is called the null hypothesis
(Ho) or sometimes called as the working hypothesis which always express the idea
of non-significance. Thus, the rejection of the null hypothesis means the acceptance
of the alternative hypothesis (Ha).

To summarize:
- The rejection of Ho implies the acceptance of Ha
- The acceptance of Ho implies the rejection of Ha

Type I and Type II Error

Type I Error (α error) – when we reject the null hypothesis (Ho) when in
fact the null hypothesis is true and the alternative hypothesis (Ha) is accepted when
in fact it is false.
Type II error (β error) – when we accept the null hypothesis (Ho) when in
fact the null hypothesis is false and the alternative hypothesis is rejected when in
fact it is true.

Level of Significance
The significance level of a test is the maximum value of the probability of
rejecting the null hypothesis Ho when in fact it is true. For tests involving statistical
decision, it is customary to use 5% or 1% level of significance.
A 5% significance level, implies that we are 95% confident that we have
made the right decision.
A 1% significance level, implies that we are 99% confident that we have
made the right decision.

One-Tailed and Two-Tailed Tests

One-tailed test signifies that the statement of alternative hypothesis is
directional, thus the rejection region is located at only one extreme of the range of
values for the test statistics.

Two-Tailed test, if the statement of alternative hypothesis is non-directional

or it signifies non-equality, the test is considered two tailed test. The level of
significance be divided into two which represents the two extremes range of values
for the test statistics.

Steps in Hypothesis Testing

1. Formulate the null hypothesis (Ho) that there is no significant difference
between items being compared. State the alternative hypothesis (Ha) which is
used in case Ho is rejected.
2. Set the level of significance.
3. Determine the test to be used. Use z-test if population standard deviation is
given, and t-test if the standard deviation given is from the samples.
4. Determine the tabular value for the test. For a z-test, use the table of the
critical values of z based on the area of the normal curve. For a t-test, compute
for the degrees of freedom then look for the tabular value from the table of t-
distribution. For a single sample, df = number of items – 1 or df = n – 1. For
two samples, df = n1 + n2 – 2
5. Compute for z or t as needed.

Statistical Tools to Test the Difference Between Two Variables

(Parametric Tests)
A. z-test
a) Sample mean compare with Population mean

Z = (x - µ)√ n
SDN
x = mean of the sample
µ = mean of the population
SDN = standard deviation of the population
n = number of items within the sample

b) Comparing two sample means

Z= ____x1 – x2____
SDN √1/n1 + 1/n2

x1 = mean of the first sample

x2 = mean of the second sample
n1 = number of items on the first sample
n2 = number of items on the second sample
SDN = standard deviation of the population

c) Comparing two sample proportions

 Z= ___p1 – p2_____
√ (p1q1)/n1 + (p2q2)/n2

p1 = proportion of the first sample

p2 = proportion of the second sample
q1 = 1 – p1
q2 = 1 – p2
n1 = number of items on the first sample
n2 = number of items on the second sample

B. t-test
a) Sample mean compare with Population mean

t = (x - µ)√ n - 1
SDn
x = mean of the sample
µ = mean of the population
SDn = standard deviation of the sample
n = number of items within the sample

b) Comparing two sample means for independent

t= x1 – x2
(n1 – 1) SD1 + (n2 – 1) SD22
2
1 + 1
√ n1 + n2 - 2 √ n1 n2

x1 and x2 = means of the first and second samples

n1 and n2 = number of items of the first and second samples
SD1 and SD2 = standard deviations of the first and second
samples

Example 1. Data from a school census show that the mean weight of college
students was 45 kilos, with standard deviation of 3 kilos. A sample of 100 college
students were found to have a mean weight of 47 kilos. Are the 100 college
students really heavier than the rest at 0.05 level of significance? Table value is
1.645.

Solution:
Ho: there is no significant difference between the mean weight of the college
students and the mean weight of the 100 samples.
Ha: The weight of the 100 samples are really heavier than the weight of the college
students.

Note: The alternative hypothesis is directional, so it entails a one – tailed test.

Level of significance = 0.05
Table value = 1.645
Standard deviation from the population = 3 kilos
Use z – test.

Z = (x - µ)√ n
SDN

= (47 – 45) √ 100

3
= 6.67

Interpretation: The computed value is greater than the tabular value. Therefore,
the null hypothesis is rejected. Accept the alternative hypothesis that the 100
samples are really heavier than the rest.

Example 2. A teacher wishes to test whether or not the Case Method of teaching is
more effective than the Traditional Method. She picks two classes of
approximately equal intelligence. She gathers a sample of 18 students to whom she
uses the Case Method and another sample of 14 students to whom she uses the
Traditional Method. After the experiment, an objective test revealed that the first
sample got a mean score of 28.60 with standard deviation of 5.90, while the second
group got a mean score of 21.70 with a standard deviation of 4.60. Based on the
result of the administered test, can we say that the Case Method is more effective
than the Traditional Method? Level of Significance is 0.05. Tabular value given is
1.697

Solution:
Ho: There is no significant difference between the case method and traditional
method of teaching.
Ha: The case method is more effective than the traditional method.
Note: The alternative hypothesis is directional, so it entails a one – tailed test.
Level of Significance = 0.05
Tabular value = 1.697

Standard deviation is taken from the samples.

 SDn1 = 5.9
SDn2 = 4.60
Mean: x1 = 28.60
x2 = 21.70

Sample Size:
n1 = 18
n2 = 14
Use: t- test, comparing two sample means for independent samples.
t= x1 – x2
(n1 – 1) SD1 + (n2 – 1) SD22
2
1 + 1
√ n1 + n2 - 2 √ n1 n2

t= 28.60 – 21.70
(18 – 1) 5.92 + (14 – 1) 4.62 1 + 1
√ 18 + 14 - 2 √ 18 14

t = 3.594

Interpretation: The computed value is greater than the tabular value. Therefore,
the null hypothesis is rejected. Accept the alternative hypothesis that the case
method is more effective than the traditional method.

Activity 3:

1. Two methods of teaching statistics are being tried by a Professor. A class of

40 students is taught by Method A and a class of 36 students is taught by Method
B. The two classes are given the same final examination. The mean scores are XA
= 78 and XB = 74. Using a 0.01 significance level, can we conclude that the
average final examination scores produced by the two methods are different if the
population standard deviation is 5? Given a table value is 2.58.

2. All freshmen in a particular school were found to have a variability in grades

expressed as a standard deviation of 3. Two samples among these freshmen, made
up of 20 and 50 students each, were found to have means of 88 and 85 respectively.
Based on their grades, is the first group really brighter than the second group at
0.01 level of significance?

3. Two kinds of feed mixes were given to 8 pigs and the average
weight gains were compared after 30 days. Test the null hypothesis that there is no
significant difference in mean gains at 0.01 level of significance.
 Feed Mixes 1 2 3 4 5 6 7 8
Mix A 60 38 39 49 62 53 42 58
Mix B 53 39 29 47 50 56 47 52

4. In a pool of television viewers, 60 out of 200 men disliked a certain

program of a TV station whereas 75 out of 300 women disliked it also. Is there a
real difference of opinion between men and women viewers?

5. A manufacturer of steel cable formerly claimed a mean breaking

strength of 55,000 kilos with a standard deviation of 500 kilos. Because of an
improved manufacturing process, the manufacturer now claims an increased
breaking strength value. A sample of 50 cables is tested and the mean breaking
strength is found to be 55,250 kilos. Does this support the manufacturer’s claim at
0.01 level of significance?

Week No. Topic: Simple Analysis of Variance (ANOVA) – The F-test

Analysis of variance is a technique in inferential statistics designed to test
whether or not more than two samples or groups are significantly different from
each other. Analysis of variance was developed by R.A. Fisher.
Elements of ANOVA
Total Variations (Vt) or the Total Sum of Squares (SSt)
This can be obtained by finding the mean of the total scores, taking the deviation
of each score from this mean, and squaring and summing these squared deviations.
It may be obtained by the use of the following equation:
SSt = ∑x2 – (∑x)2/N
Variation Between (Vb) or the Between Sum-of-Squares (SSb)
The sum of the squares between the various groups can be found by taking the
mean of each group, getting its deviation from the total mean, squaring this
deviation, and then multiplying each of these by the number of individuals in each
group as follows:
SSb = ∑ (x – xt)2 n
Or SSb = ∑ (∑x)2 - (∑xT)2
n N
Variation Within (Vw) or the Within Sum-of-Squares (SSw)
To obtain the within sum-of-squares we could find the sum-of-squares of each
group. The within sum-of-squares added to the between sum-of-squares should
total the total sum-of-squares. It follows then, that the within sum-of-squares can
be obtained directly by subtracting the between sum-of-squares from the total sum-
of-squares.
SSw = SSW1 + SSW2 + SSW3 ……SSWn
SSW1 = ∑x2 – (∑x)2
SSW + SSB = SST
Degrees of Freedom
df for total groups = number of cases in total (N) minus 1
df for groups between = number of groups (k) minus 1
df for groups within = sum of the number of cases within each
sub-groups (n) minus 1. Thus dfw = (n1 – 1) + (n2 – 1) + (nk – 1)
The F- Test
The analysis of variance is evaluated by making the following F test.
F = Mean square for between groups
Mean square for within groups
Decision:
A. If the computed value of F is less than the tabular value the null
hypothesis (Ho) is accepted, which means that there is no significant difference
between the variables being compared.
B. If the computed value of F is greater than the tabular value the null
hypothesis (Ho) is rejected. Which means that difference exists between the
variables being compared.
If difference is existing between the variables compared, then determine where
the difference exists between the variables. It is where tests after the F test should
be computed using the Scheffe’ technique.
F = (x1 – x2)2 N1N2
S2w (N1 + N2)
Interpretation of the result for inferential statistics using Sig value.
If the computed Sig. value is less than the signified error α (0.05 or 0.01),
then significant difference exists. Therefore, Reject the Null Hypothesis, stating
that there is no significant difference between the variables being compared.
If the computed Sig. Value is greater than the signified error α, then no
significant difference exists. Therefore, Accept the Null hypothesis, stating that
there is no significant difference between the variables being compared.
Simple Method for Calculation Analysis of Variance
A. For Equal Number of Observations
Variations Degrees of Mean Squares F
Freedom
Between Treatments S2B = VB F = S2B
VB = b ∑(xj – x)2 a-1 a-1 S2W
Within Treatments S2W = VW
VW = ∑ (x – xj)2 a (b – 1) a (b – 1)
Total Variations
VT = ∑ (x – x)2 ab - 1

B. For Unequal Number of Observations

Variations Degrees of Mean Squares F
Freedom
Between Treatments S2B = VB F = S2B
VB = nj ∑(xj – x)2 a-1 a-1 S2W
Within Treatments S2W = VW
VW = ∑ (x – xj)2 n–a a (b – 1)
Total Variations
VT = ∑ (x – x)2 n-1
Example 1. From the given data, determine if difference exists between the groups
A, B, and C. If the results showed significant differences, apply Scheffe’ technique
to find out where the differences are.
Group A Group B Group C
12 18 6
18 17 4
16 16 14
8 18 4
6 12 6
12 17 12
10 10 14

Solution:
Ho: There are no significant differences on the mean scores of Groups A, B,
and C.
Ha: Significant differences exist between the mean scores of Groups A, B,
and C.
Level of Significance = 0.05
Table value = 3.55
Group A Group B Group C Group A Group B Group C
X X X X2 X2 X2
12 18 6 144 324 36
18 17 4 324 289 16
16 16 14 256 256 196
8 18 4 64 324 16
6 12 6 36 144 36
12 17 12 144 289 144
10 10 14 100 100 196
82 108 60 1068 1726 640

Xj = 11.71 15.43 8.57

∑X = 82 + 108 + 60 = 250 ∑x2 = 1068 + 1726 + 640 = 3434

X = 11.9 N = 21
SSt = ∑x2 – (∑x)2/N
= (122 + 182 + 162 +…………+ 122 + 142) – 2502/21 = 457.80
or
SSt = (1068 + 1726 + 640) – (82 + 108 + 60)2 = 457.80
21
SSb = ∑ n (x – xt) 2

= 7(11.71 – 11.9)2 + 7(15.43 – 11.9)2 + 7 (8.57 – 11.9)2 = 165.10

or
SSb = ∑ (∑x)2 - (∑xT)2
n N
= (822/7 + 1082/7 + 602/7) – 2502/21 = 165.0
SSw = SSW1 + SSW2 + SSW3 ……SSWn
SSW1 = ∑x2 – (∑x)2/n
SSWA = 1068 – 822/7 = 107.4
SSWB = 1726 – 1082/7 = 59.7
SSWC = 640 – 602/7 = 125.7
SSw = 107.4 + 59.7 + 125.7 = 292.8
dfT = 21 – 1 = 20
dfB = 3 – 1 = 2
dfW = (7 -1) + (7 – 1) + (7 – 1) = 18

ANOVA Table:
Source of Sum-of- Degrees of Mean Compute
Variation Squares freedom Square dF
Between 165 2 82.5 5.06
groups
Within 292.8 18 16.3
Groups
Total 457.8 20

Interpretation:
The Computed F value (5.06) is greater than the tabular value (3.55). This
means that the null hypothesis is rejected, that there are no significant differences
on the scores of the groups A, B, and C. Therefore, the alternative hypothesis is
accepted, that differences are existing between the given scores of groups A, B, and
C.
Test after F test
F = (x1 – x2)2 n1n2
S2w (n1 + n2)
A Vs. B; A Vs. C; B Vs. C.
FAB = (11.71 – 15.43)2 (49) = 2.97
16.3 (14)
FAC = (11.71 – 8.57)2 (49) = 2.12
16.3 (14)
FBC = (15.43 –8.57)2 (49) = 10.10
16.3 (14)
Note: The value of F greater than the tabular value of 3.55 is 10.10 and that is for
Groups B and C. So, the difference exists between this group. Investigating the
scores of the two groups it is noted that the scores of group C is much lesser
compare to the scores of group B.
Activity 4:
1. The following represent the test scores of three groups of students.
a) Do the means of the three groups differ significantly? Level of significance
is 0.01.
b) If so, apply the Scheffe’ test to find out where the differences are.

Group A Group B Group C

10 4 5
9 6 2
8 4 6
7 8 7
12 6 10
10 4 2
11 5 1
12 7 4
9 10 3
8 11 4

2. A company wishes to purchase one of the five different machines A, B,

C, D, E. In an experiment designed to decide whether there is a difference in
performance of the machines, five experienced operators each work on the
machines for equal times. The table shows the number of units produced. Test the
hypothesis that there is no difference between the machines at the a) 0.05, b) 0.01
level of significance.
A 68 72 75 42 53
B 72 52 63 55 48
C 60 82 65 77 75
D 48 61 57 64 50
E 64 65 70 68 53

3. The Table shows the lifetimes in hours of samples from three different
types of television tubes manufactured by a company. Test at 0.05 and 0.01 level
of significance whether there is a difference in the three types of television tubes.
Sample 1 407 411 409
Sample 2 404 406 408 405 402
Sample 3 410 408 406 408