INEN 30063

STATISTICAL ANALYSIS FOR IE -2

Continuation of
LESSON 1 :REVIEW OF
SIMPLE TESTS OF HYPOTHESIS

October 12, 2023

-Engr. R. Ilagan-
Hypothesis is a statement or tentative theory or assertion or
conjecture which aims to explain facts about the real world.

Two Types of Statistical Hypotheses

1. A null hypothesis expresses the idea of non-significance of

difference or the idea of no relationship.

2. An alternative hypothesis is the opposite of the null hypothesis.

It generally represents the hypothetical statement that the
researcher wants to prove or claim the manufacturer eager to
substantiate. Alternative hypotheses are classified as either
non-directional or directional hypotheses.
Alternative Hypotheses

1. A non-directional hypothesis (Ha) is one which asserts that one

value is different from another (or others). More specifically, it is
an assertion that there is a significant difference or significant
relationship between two statistical measures (or among three
or more summary measures). It implies non-equality (≠).

2. A directional hypothesis (H1), sometimes called predictive

hypothesis, is an assertion that one measure is less than (or
greater than) another measure of similar nature. Mathematically,
a non-directional hypothesis makes use of the “not equal to” (≠)
sign, while a directional hypothesis involves one of the order
relations, “less than” (<) or “greater than” (>). Non-directional
hypotheses are also called two-sided hypotheses and
directional hypotheses are also known as one-sided hypotheses
Illustrations

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

professor. A class of 40 students by the method A and a class
of 36 is taught by method B. The two classes are given the
same final examination. The mean scores are 78 and 74,
respectively. Using a 0.05 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?

Remark:
The word different implies non-equality (≠). Thus, the problem
necessitates a non-directional hypothesis (Ha) or just simply
called alternative hypothesis.
Illustrations

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

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?

Remark:
The word brighter than implies inequality (>). Thus, the problem
necessitates a directional hypothesis (Ha) or just simply called
predictive hypothesis.
 Simple Tests of Hypothesis

A test of hypothesis is a procedure used to substantiate or

invalidate a claim which is stated as a null hypothesis. The
rejection of null hypothesis is to conclude that the hypothesis
is false, and therefore, the alternative hypothesis is accepted.
The acceptance of the null hypothesis merely implies that
there is no sufficient statistical evidence to believe otherwise.
The goal of hypothesis testing is not to question the computed
value of the sample statistic but to make a judgment about the
difference or existence of relationship between the sample
statistics and a hypothesized population parameter. It enables
a researcher to generalize something about the population
from studying a relatively small sample. In many instances, a
researcher can only rely on the information provided by a part
of the population
Simple Tests of Hypothesis
A. Comparison of Means of Two Large/Small Independent Samples When Population
Variances (Standard Deviations) Are Known
( x - y ) - d0 ( x - y ) - d0
z= when  x   y ; z = when  =  x =  y
x y
2 2 1 1
+  +
n m n m
B. Comparison of Means of Two Large Samples When Population Variances (Standard
Deviations) Are Unknown
( x - y ) - d0
z=
2
s 2x s y
+
n m
C. Comparison of Means of Two Independent Small Samples When Population Variances
(Standard Deviations) are Unknown
Assumption :  x =  y A ssumption :  x   y
( x - y ) - d0 ( x - y ) - d0
t= t=
1 1 2
s + s 2x s y
n m +
n m
(n - 1)s 2x + (m - 1)s 2y
sp =
n+m-2
sp = pooled standard deviation df = n + m - 2
D. Comparison of Means of Two Correlated/Dependent Small Samples When Population
Variances (Standard Deviations) are Unknown
D
t computed = where D = difference of paired scores; n = no. of pairs
n D - (D)
2 2

n -1
dfcorrelated samples = n - 1 (Pangan, et.al., 1996) ( Basilio, Faith B., et.al., 2003)
When To Use A Particular Test
When a sample is large, it is appropriate to conduct a z-test because the
distribution of sample means approaches a normal distribution.
However, when the sample is relatively small, the distribution of sample
means is best matched by the t-distribution, not the z-distribution. The
formula for computing the t score is essentially the same as that for
computing the z score:
A. z-test or t-test
1. z-test is used when the population standard deviation is known,
or in the absence of population standard deviation, the sample
size is equal to or greater than thirty.
2. t-test is used when the population standard deviation is
unknown and the sample size is less than thirty.
B. One-Tailed Test or Two-Tailed Test
1. Non-directional alternative hypothesis indicates the use of a
two-tailed test.
2. Directional predictive hypothesis indicates the use of a one-
tailed test.
 Illustrations

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

professor. A class of 40 students by the method A and a class
of 36 is taught by method B. The two classes are given the
same final examination. The mean scores are 78 and 74,
respectively. Using a 0.05 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?

Remark:
A. Comparison of Means of Two Large/Small Independent
Samples When Population Variances (Standard Deviations)
Are Known
( x - y ) - d0 ( x - y ) - d0
z= when  x   y ; z= when  =  x =  y
 x2  y2 
1 1
+
+ n m
n m
Illustrations

2. The average grade of 50 senior students in statistics is 85 with

a standard deviation of 10.2, while a group of 60 senior
students got an average of 80 with standard deviation of 8.9.
Can the difference in the mean grade be attributed to chance,
using 0.05 level of significance?

Remark:
B. Comparison of Means of Two Large Samples When
Population Variances (Standard Deviations) Are Unknown

( x - y ) - d0
z=
s 2x s 2y
+
n m
Illustrations

3. Two groups, X and Y of freshman students, 28 in each group, are paired for age and
score on form A of the Otis Group Intelligence Scale. Three weeks later, both groups
are given form B of the same test. Before the second test, Group X, the experimental
group, is praised for its performance on the first test and urged to try to score better.
Group Y, the control group, was given the second test without comment. Will the
incentive (praise) cause the final scores of Group X and Group Y to differ
significantly? Use alpha equal to 0.10.
Group X Group Y
x = 88.63 y = 83.24
sx = 24.36 sy = 21.62
Remark:
C. Comparison of Means of Two Independent Small Samples When Population
Variances (Standard Deviations) are Unknown
Illustrations

4. A group of employees was given an attitude test on E-VAT; then, they were shown a
film favorable to implementation of E-VAT. The attitude test was then re-
administered. Is there a significant difference between their attitudes before and
after they have seen the film? Use α = 0.10. The data are shown below.
X Y
15 25
17 19
19 23
23 27
23 30
21 20
19 23
18 21
9 17
8 17
Remark:
D. Comparison of Means of Two Correlated/Dependent Small Samples When
Population Variances (Standard Deviations) are Unknown
D
t computed =
n D 2 - (D )
2

n-1
dfcorrelated samples = n - 1
DECIDING ON WHETHER TO ACCEPT OR REJECT THE NULL HYPOTHESIS

A. Using the Normal Distribution

1. If the computed z-value (or t-value) falls within the
acceptance region, accept the null hypothesis.
2. If the computed z-value (or t-value) falls within the
rejection region, reject the null hypothesis.

B. One-Tailed Test or Two-Tailed Test

1. If the absolute value of computed z-value (or t-value) is
less than or equal to the absolute value of the critical z-
value (or t-value), accept the null hypothesis.
2. If the absolute value of computed z-value (or t-value) is
greater than the absolute value of the critical z-value (or t-
value), reject the null hypothesis.
t- test table
Illustrations
1. Two methods of teaching statistics are being tried by a professor. A class of 40 students by the method A and a class of
36 is taught by method B. The two classes are given the same final examination. The mean scores are 78 and 74,
respectively. Using a 0.05 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?
1. Given:
Population x Sample y Sample α = 0.05
σ=5 nx = 40 ny = 36
x = 78 y = 74
Type of Problem: Comparison of means of two large/small samples
when population variance (standard deviation) is known
( x - y ) - d0 ( x - y ) - d0
z= when  x   y ; z = when  =  x =  y
x2
 2

1
+
1
+ y
n m n m
(a) Hypotheses
H0: There is no significant difference between the mean scores of
sample students taught by methods A and B.
μx = μy (mathematical model)
Ha: There is significant difference between the mean scores of
sample students taught by methods A and B.
μx ≠ μy
(b) Level of Significance: α = 0.05
(c) Test Statistic
Population standard deviation is given, Z-test is applied.
Non-directional alternative hypothesis, two-tailed test is applied.
( x - y ) - d0 (78 - 74) - 0
zcomputed = = = 3.48
1 1 1 1
 + (5) +
n m 40 36
zcritical = ± 1.96 @ α = 0.05 and two-tailed test
(d) Decision: Since │ Zcomputed │ > │Zcritical │ reject Ho, accept Ha.
(e) Conclusion: There is significant difference between the mean scores of
sample students taught by methods A and B. μx ≠ μy
Illustrations
2. The average grade of 50 senior students in statistics is 85 with a standard deviation of 10.2, while a group of 60 senior
students got an average of 80 with standard deviation of 8.9. Can the difference in the mean grade be attributed to
chance, using 0.05 level of significance?

2. Given:
x Sample y Sample α = 0.05
nx = 50 ny = 60
x = 85 y = 80
Sx = 10.2 Sy = 8.9
Type of Problem: Comparison of means of two large samples when
population variance (standard deviation) is unknown
( x - y ) - d0
z=
s 2x s 2y
+
n m
(a) Hypotheses
H0: There is no significant difference between the average grades of
50 sample senior students and 60 sample senior students.
μx = μy (mathematical model)
Ha: There is significant difference between the average grades of
50 sample senior students and 60 sample senior students.
μx ≠ μy
(b) Level of Significance: α = 0.05
(c) Test Statistic
Population standard deviation is not given and n > 30, Z-test is applied.
Non-directional alternative hypothesis, two-tailed test is applied.
( x - y ) - d0 (85 - 80) - 0
zcomputed = = = 2.71
2 2 2 2
sx s (10.2) (8.9)
+ y +
n m 50 60
z critical = ± 1.96 @ α = 0.05 and two-tailed test
(d) Decision: Since │ Zcomputed │ > │Zcritical │ reject Ho, accept Ha.
(e) Conclusion: There is significant difference between the average grades
of 50 sample senior students and 60 sample senior students. μx ≠ μy
 Illustrations
3. Two groups, X and Y of freshman students, 28 in each group, are paired for age and score on form A of the Otis Group Intelligence
Scale. Three weeks later, both groups are given form B of the same test. Before the second test, Group X, the experimental
group, is praised for its performance on the first test and urged to try to score better. Group Y, the control group, was given the
second test without comment. Will the incentive (praise) cause the final scores of Group X and Group Y to differ significantly?
Use alpha equal to 0.10. Group X ( = 88.63; sx = 24.36) and Group Y ( = 83.24; sy = 21.62)
3. Given:
x Sample y Sample α = 0.10
nx = 28 ny = 28
x = 88.63 y = 83.24
Sx = 24.36 Sy = 21.62
Type of Problem: Comparison of means of two small independent samples
when population variance (standard deviation) is unknown
( x - y ) - d0 (n - 1)s2x + (m - 1)s2y
t= s= df = n + m - 2
1 1 n+m-2
s +
n m
(a) Hypotheses
H0: There is no significant difference between the mean scores of group x
and y of 28 sample students each.
μx = μy (mathematical model)
Ha: There is significant difference between the mean scores of group x and
y of 28 sample students each.
μx ≠ μy
(b) Level of Significance: α = 0.10
(c) Test Statistic
Population standard deviation is not given and n < 30, T-test is applied.
Non-directional alternative hypothesis, two-tailed test is applied.
(n - 1)s2x + (m - 1)s2y (28 - 1)(24.36)2 + (28 - 1)(21.62)2
s= = = 23.03
n+m-2 28 + 28 - 2
( x - y ) - d0 (88.63 - 83.24) - 0
t computed = = = 0.88
1 1 1 1
s + 23.03 +
n m 28 28
df = n + m - 2 = 28 + 28 - 2 = 54
t critical = ± 1.674 @ α/2 = 0.05 and two-tailed test
(d) Decision: Since │ T computed │ < │Tcritical │ accept Ho.
(e) Conclusion: There is no significant difference between the mean scores of
group x and y of 28 sample students each. μx = μy
 Illustrations

4. A group of employees was given an attitude test on E-VAT; then, they were
shown a film favorable to implementation of E-VAT. The attitude test was then
re-administered. Is there a significant difference between their attitudes
before and after they have seen the film? Use α = 0.10. The data are shown
below.

4. Given:
X Y D (y-x) D squared
15 25 10 100
17 19 2 4
19 23 4 16
23 27 4 16
23 30 7 49
21 20 -1 1
19 23 4 16
18 21 3 9
9 17 8 64
8 17 9 81
50 356
 Illustrations
4. A group of employees was given an attitude test on E-VAT; then, they were shown a
film favorable to implementation of E-VAT. The attitude test was then re-administered.
Is there a significant difference between their attitudes before and after they have seen
the film? Use α = 0.10. The data are shown below.
Type of Problem: Comparison of means of two small correlated samples when
population variance (standard deviation) is unknown
D
t computed =
n D2 - (D )
2

n -1
dfcorrelated samples = n - 1
(a) Hypotheses
H0: There is no significant difference between the attitudes of the
employees before and after they have seen the film. μx = μy
(mathematical model)
Ha: There is significant difference between the attitudes of the
employees before and after they have seen the film. μx ≠ μy
(b) Level of Significance: α = 0.10
(c) Test Statistic
Population standard deviation is not given and n < 30, T-test is applied.
Non-directional alternative hypothesis, two-tailed test is applied.
D 50
t computed = = = 4.607
n D - (D) 10 (356) - (50)
2 2 2

n -1 10 - 1
dfcorrelated samples = n - 1 = 10 - 1 = 9
t critical = ± 1.833 @ α/2 = 0.05, df = 9 and two-tailed test
(d) Decision: Since │ Tcomputed │ > │Tcritical │ reject Ho, accept Ha.
(e) Conclusion: There is significant difference between the attitudes of the
employees before and after they have seen the film. μx ≠ μy
 EXERCISES -HYPOTHESIS TESTING

EXERCISES. FOLLOW THE STEPS IN HYPOTHESIS TESTING.

1.A data from Golden State University in the region shows that the mean
grade of Engineering students is 89 with a standard deviation of 4. A sample
of 150 students was to have a mean grade of 91. Are the engineering students
performing better than the rest using 0.01 level of significance .?.
2.According to the Civil Code of the Philippines , the marriageable age of
Filipinos is 18. A researcher found out that the average of Filipinos who got
married is 23 years old. A random sample 0f 29 married couples were taken
and found out to have an average age of 21 with a standard deviation of 1.23.
Is there a reason to believe that the sample is significantly younger than the
others in getting married at 1% level of significance?.
3. Does an average box of Brand X full cream milk contains more than 350
grams of milk? A random sample of 30 box showed a mean of 357. The
company has specified a standard deviation to be 10 grams . Test at the α=
0.05 level.
 EXERCISES -HYPOTHESIS TESTING

4. A random sample of 25 brand M corned beef showed a

mean net weight of 190 grams while a brand N corned beef
showed a mean net weight of 188 grams. The population
standard deviation is given to be 3.5 grams. Is there a
difference in the mean net weight of the two brands of
corned beef ?. Set the level of significance at 5%.
5. A test in statistics was given to the groups of Business
Administration students . 31 students majoring in Finance
got an average score of 78 with a standard deviation of 3.5 ,
while 31 students majoring in Management got an average
score of 81 with a Standard deviation of 2.5 . Do the
performance of the two groups differ?. Assume that the
distribution normal . Use alpha at 1%.