You are on page 1of 9

Inferential Statistics

⮚ Includes all the techniques used


in analysing the sample data that
will lead to generalizations about
a population from which the
sample came from.

Correlation Analysis
⮚ Many research studies in the
field of education, psychology,
sociology, industry, business and
medicine deal with relationships
The Pearson Product Moment
between two or more variables.
Coefficient Correlation, r
Statistics provide us with
techniques and tools necessary
⮚ It is used to analyze if a
in making such studies. One of
relationship exists between two
these is correlation analysis,
variables (measured in the
which is used to measure the
interval or ratio scale) say
relationship that exists between
variable x and y.
two variables.
⮚ It was developed by Karl Pearson
that is why the correlation
Types of correlation
coefficient is sometimes called
I. Positive Correlation
"Pearson's r." The formula is
⮚ Eg: Height & weight
defined by:

II. Negative Correlation: Basic properties of r

⮚ The range of the correlation


coefficient is -1 and +1.
⮚ If the value of the coefficient is
close to -1.00, it represents a
⮚ Eg. Price & quantity demanded negative correlation while a value
of +1.00 represents a positive
correlation.
⮚ If the value is equal to 0.00, it
means that there is no relation
between the variables.
The degree of relationship can be
interpreted through the use of the
range of values.

The Pearson Product Moment

Coefficient Correlation, r
SPEARMAN RHO
⮚ Sometimes, when numbers are
large, it will be easier to use
another statistical tool for
computation of correlation.
Spearman rho (ρ) can be used,
but data should be changed first
to ordinal data.
Computational Formula for Spearman
Rho:

Where:
⮚ ρ = the degree of relationship
between the two independent
variables
⮚ D = difference between ranks
of the variables
⮚ N = number of cases
⮚ The alternative hypothesis,
denoted by 𝐻𝑎, is the operational
statement of the theory that the
researcher believes to be true
and wishes to prove and is a
contradiction of the null
hypothesis.
Null vs Alternative
• These two hypotheses must be
non-overlapping statements.
• The null hypothesis is usually
expressed as a hypothesis of
Applying the formula: equality or a hypothesis of no
effect.
• On the other hand, common type
of alternative hypothesis is a
composite expression about the
direction of the parameter from
the hypothesized value.
Test of Hypothesis
⮚ A test statistic is a statistic whose
Interpretation: value is calculated from sample
data, which be the basis for
⮚ There exist a low positive
deciding whether to reject 𝐻0 or
correlation between the height
of the father and the height of not in a test of hypothesis.
the eldest son, that is, a tall ⮚ The critical region is the set of
father does not necessarily values of the test statistic for
mean he will have a tall son, which we reject the null
likewise a short father does not hypothesis. The acceptance
necessarily mean he would region is the set of values of the
have a short son. test statistic for which we do not
reject the null hypothesis.

Statistical Hypothesis - Testing of These two regions are separated by the


hypothesis critical value of the set statistic

⮚ A null hypothesis, denoted by 𝐻0, Type of Tests


is a statistical hypothesis which 1. One- tailed test - directional
the researcher doubts to be true.
a. left- tailed test or not the null hypothesis is
rejected.
❖ H1 : parameter < value
⮚ We reject the null hypothesis if
Decision: Reject Ho if test statistics is
the test statistic is greater than to
less than the critical value
the critical value.
b. right –tailed test
❖ H1: parameter > value
Decision: Reject Ho if test statistics is
greater than the critical value
2. Two – tailed test
❖ H1 : parameter “ not equal” (≠)
Decision: Reject H1 if the test statistics
is between the criticall value specified

Critical Value
⮚ The critical value or the tabular
value for the hypothesis test is a
threshold to which the value of
the test statistic in a sample is
compared to determine whether
extreme than the realized
value computed from the
sample data, given that the
null hypothesis is true.
As a rule, if the p-value is less than the
level of significance, then we do not
reject the null hypothesis. On the other
hand, if the p-value is greater than or
equal to the level of significance, then
we reject the null hypothesis.
Steps in Hypothesis Testing
Types of Error
1. State the null and alternative
⮚ The Type I error is the error hypotheses.
committed when we decide to
2. Choose the level of significance
reject the null hypothesis when in
reality the null hypothesis is true. 3. Determine the appropriate
⮚ The Type II error is the error statistical technique and
committed when we decide not to corresponding test statistic to
reject the null hypothesis when in use.
reality the null hypothesis is false.
4. Perform the computation.
The Level of Significance Compare the computed value
with the critical value (others use
⮚ The level of significance,
the p-value instead)
denoted by α, is the maximum
probability of committing a type I 5. Make the decision rule (Reject
error that the researcher is willing the null hypothesis or failed to
to commit.Very frequently used reject it).
are the .10, .05 and .01 level of
Decision Rule
significance.
⮚ Reject 𝐻0 if the value of the test
⮚ Note: 0.05 level of significance
implies that we are willing to statistic falls in the region of
commit an error of 5% therefore a rejection (that is, test statistics is
confidence level of 95%. greater than the critical value.)
⮚ Reject 𝐻0 if the p-value is less
P-value than or equal to the level of
significance.
⮚ The p-value is the probability
of selecting a sample whose
computed value for the test
statistic is equal or more
⮚ A parametric test applied to one
group of samples.
⮚ It can be used in evaluation of a
certain program or treatment.

⮚ It is applied when the mean


before and the mean after are
being compared.

Test statistic

T-test for Independent Samples


(unpaired)
⮚ Used when we compare the
means of two independent
groups.

⮚ Used when the sample is used


and not the population

T-test for Dependent Samples


(paired)
A group of researchers claim that
the mean number of children born
to urban women is less than the
mean number of children born to
rural women.
Let:
µ1 = mean of children born to
urban women
µ2 = mean number of children born
to rural women
Example 3:
Prior to the new safety program of the
MMDA the average number of road
accidents per day is 5.5. To determine if
the safety program has been effective in
reducing the average number of
accidents per day, a random sample of
30 days is taken and recorded.
Ho : µ = 5.5
H1 : µ < 5.5
The following examples show how to
formulate the different kinds of
hypothesis:
Procedure for testing hypothesis
Example 1:
Step 1. State the hypotheses
A Biostatistics professor claims that
Step 2. Find the critical value from the
using a module will enhance the
tables
performance of his students. In the past,
the average grade of his students was Step 3. Compute the test value using t-
2.25. test and or z-test
Null hypothesis: Ho :µ ≤ 2.25 Step 4. Make the decision to reject or
accept the null hypothesis
Alternative hypothesis: H1 : µ > 2. 25
Step 5. Summarize result

Example 2:
Small Sample Mean Test Step 3. Compute the test value
⮚ A different test called the t- test is t= Ⴟ-µ
used when the sample size is
less than 30, and the sample s/√n
standard deviation is given. t = 725 - 620
The formula for 22 / √ 20
the t –test is :
t = 105
t= Ⴟ-µ
22/ 4.47
s/√n
t = 105
The degree of freedom is computed if
t-test is used. 4.92

df=n–1 = 21.34

Step 4. Make the decision


Example: Reject the null hypothesis since 21.34 is
greater than 2.54
A certain food company claims that the
average calorie content of its ready- to – Step 5. Summarize the result
eat meals is 620, the standard deviation
is 22. A researcher tested 20 meals and There is enough evidence to reject the
found out that the average number of claim of the food company that the
calories was 725. Is there an evidence average calorie content of their
to reject the claim at α = 0.01, assuming ready-to-eat meal is 620.
the variable is normally distributed?
Step 1. State the hypothesis
Ho : µ ≤ 620 (claim) (one- tailed test)
H1 : µ > 620
Step 2. Find the critical value
At α = 0.01 and df = 20 – 1 = 19, the
critical value is ± 2.54

You might also like