HYPOTHESIS

TESTING
Descriptive Versus Inferential
Descriptive statistics is required to
demonstrate that there is a difference
between the two groups, and
inferential statistics is needed to show
that if the experiment were repeated,
it would be likely that the difference
would be in the same direction and
somehow you can create a conclusion
out of that analysis from a sample.
Parametric Versus Nonparametric
If the research data gathered is “smooth” in its
distribution, we call it normally distributed, and if it
follows its parameters there are statistical procedures
used to infer the research when these two conditions
are met. These procedures are called parametric
statistics. If the statistical data don’t meet the
requirements under normal distribution, we call it
nonparametric statistics.
Levels of Measurement
1.Nominal Scale
• This is characterized by data that consists of names, labels, or categories
only. The data cannot be arranged in an ordering scheme.
• When numerical values or symbols are used to classify an object, person, or
characteristics to identify groups to which various objects, persons, or
characteristics belong. They are used merely for classification or
identification purposes.
2.Ordinal Scale
• This is a level of measurement which contains the properties of the nominal
level, but also gives arrangement of members of particular grouped in
order, rank, or ordered in some low-to-high manner. Inequalities between
data values cannot be determined.
3.Interval Scale
• It contains the properties of the ordinal level but the distances between any
two numbers on the scale are of known sizes.
• It is characterized by a common and constant unit of measurement but the
units are arbitrary.
• The number zero does not imply the absence of the characteristics under
consideration (the zero point is arbitrary).
4.Ratio Scale
• It contains the properties of the interval level but it has a true zero point, that
is, the number zero indicates the absence of the characteristic under
consideration.
• This refers to a property defined by an operation which permits making of
statements of equality of ratios in addition to statements of sameness of
differences, greater than or less than, and equality or inequality of differences.
• The strongest level of measurement.
Hypothesis Testing is the process of
inferring a sample whether to accept or
reject a certain statement about a
population. Hypothesis is defined as a
proposition placed under examination,
which cannot be accepted as true unless
proven and tested statistically. A working
hypothesis is an educated guess that a
researcher made about the chosen sample
used in the study.
Two Kinds of Hypothesis
1.Null Hypothesis (Ho) – expresses the idea of
no existence of relationship or difference
between the variables under the study. It is
the hypothesis to be tested.
2.Alternative Hypothesis (Ha) or Research
Hypothesis – it states the existence of
difference or relationship between variables.
It is the hypothesis that the researcher tries
to establish.
The hypothesis is tested either directional or
not:
1.Directional Test of Hypothesis – is
considered if the direction of the difference
is stated. This test makes use of the one-
tailed test.
2.Non-directional Test of Hypothesis – is
considered if the direction or the nature of
the difference is not stated. This test makes
use of the two-tailed test.
6-Step Model of Hypothesis Testing

1.Formulate the null hypothesis and

the alternative hypothesis
2.Set the level of significance (α)
• The level of significance is the
maximum probability of rejecting a
null hypothesis.
There are 2 ways of making
incorrect decisions:
a.Type I error – when the null
hypothesis is true and yet
rejecting it.
b.Type II error – when the null
hypothesis is false and yet
accepting it.
3.Determine the appropriate statistical test to be
used.
a.Use z-test if
✓The population standard deviation is
known
✓The sample size is large (n ≥ 30)
b.Use t-test if
✓The standard deviation is estimated from
the sample
✓The sample size is small (n < 30)
4.Determine the critical value or the tabular
value for the test
For t-test, it is a must to compute the degrees
of freedom (df). The number of degrees of
freedom is equal to the number of the
remaining values in a data set that are free to
vary after one or more values have been
deducted.
For 1 sample, df = n – 1
For 2 samples, df = n1 + n2 – 2
Two-tailed Test
It has a critical region consists of two parts.
The total probability for both tails is α and the
probability of each tail is α/2.
One-tailed Test
Its critical region consists of only one part, which is
either left or right tail. The probability with the
selected tail is α.
If the alternative hypothesis contains
the greater than symbol (>), then the
rejection region is on the right tail of
the curve. Otherwise, the rejection falls
on the left side of the curve.
5.Compute for statistical test
The selection of the statistical test to be used in
analysing the data depends on the level of
measurement of the data.
Formulas for computing the value of statistical test
z-test for 1 sample mean
𝑥ҧ − 𝜇 𝑛
𝑧=
𝜎
z-test for 2 sample means
𝑥ҧ1 − 𝑥ҧ2
𝑧=
1 1
𝜎 +
𝑛1 𝑛2
z-test for two sample proportions
𝑃1 − 𝑃2
𝑧=
𝑃1 𝑞1 𝑃2 𝑞2
+
𝑛1 𝑛2

z-test for 2 populations

• When σ1 and σ2 are given: • When s1 and s2 are given:
𝑥ҧ 1 − 𝑥ҧ 2 𝑥1ҧ − 𝑥ҧ2
𝑧= 𝑧=
𝜎1 2 𝜎2 2
+ 𝑠1 2 𝑠2 2
𝑛1 𝑛2 +
𝑛1 𝑛2
t-test for 1 sample mean
𝑥ҧ − 𝜇 𝑛
𝑡=
𝑠
t-test for 2 sample means
𝑥1ҧ − 𝑥ҧ2
𝑡=
𝑛1 − 1 𝑠1 2 + 𝑛2 − 1 𝑠2 2 1 1
+
𝑛1 + 𝑛2 − 2 𝑛1 𝑛2
6.Compare the computed value with the
tabular value. State your conclusion and
interpret your result.
To determine the significance of the
computed value:
a.If the absolute computed value is less than
the absolute tabular value, Ho is accepted.
b.Otherwise, Ho is rejected and then Ha is
accepted.
Example no. 1:
A data from ABC university in the region shows that the mean grade of
nursing students is 89, with a standard deviation of 4. A sample of 150 nursing
students was found to have a mean grade of 91. Are the 150 nursing students
performing better than the rest using 0.01 level of significance?
Sol’n:

1. Ho: The 150 nursing students are not really performing better academically
than the rest.
Ha: the 150 nursing students are really performing better academically
than the rest.
2. α = 0.01
3. z-test for 1 sample [is to be used] 6. Therefore, Ho is rejected.
4. Tabular value is ±2.33 6.12 > ±2.33
ҧ
𝑥−𝜇 𝑛 91 −89 150
5. z = = Thus, the 150 nursing students
𝜎 4
performing better academically than
𝑧 = 6.12 the rest.
 Sol’n:
Example no. 2:
1. Ho: There is no significant difference
A survey of the most between the two proportions.
requested song in a radio Ha: There is a significant difference
program in the city shows that between the two
95 out of 150 men like the song proportions.
requested while 110 out of 200 2. α = 0.05
3. z-test for 2 sample proportions
women like also the song
4. Tabular value is ±1.96
requested. A researcher wants 95
5. 𝑃1 = = 0.63
to decide whether the 150
110
difference between the two 𝑃2 = = 0.55
200
𝑃1 −𝑃2 0.63 −0.55
sample proportions is 𝑧 = 𝑃𝑞 𝑃𝑞 =
1 1+ 2 2 0.63 1−0.63 0.55 1−0.55
significant or not at 5% level of 𝑛1 𝑛2 150
+
200
significance. 𝑧 = 1.51
6. Therefore, Ho is accepted.
1.51 < ± 1.96
 Reader is alone Reader is not alone
Example #3:
68 70
Dr. Amalang records
74 76
the average number of
42 49
errors an individual 50 48
makes in reading aloud. 52 56
Each individual is tested 49 54
once. The scores for each 60 65

participant are given in 56 56

the table. Test the 69 65

following data by using 5% 65 67

level of significance. 63 63

72 65
Solution:
1. Ho: There is no significant difference between Σ𝑥1 720 σ 𝑥 − 𝑥ഥ1 2
𝑥ҧ1 = = = 60 𝑠1 = = 10.198
the two samples. 𝑛1 12 𝑛1 − 1
Ha: There is a significant difference between the two
samples.
2. α = 0.05 Σ𝑥2 734 σ 𝑥 − 𝑥ഥ2 2
3. t−test for 2 sample means [is to be used.] 𝑥ҧ2 = = = 61.17 𝑠2 = = 8.558
𝑛2 12 𝑛2 − 1
4. df = n1 + n2 − 2 = 12 + 12 − 2 = 22
Tabular value is ± 2.074
𝑥ҧ1 − 𝑥ҧ2
𝑡=
𝑛1 − 1 𝑠1 2 + 𝑛2 − 1 𝑠2 2 1 1
𝑛1 + 𝑛2 − 2 +
𝑛1 𝑛2

60.0 − 61.17
𝑡=
12 − 1 10.198 2 + 12 − 1 8.558 2 1 1
+
12 + 12 − 2 12 12

𝑡 = −0.30357

−0.30357 < ±2.074 ; Ho is accepted