Theorem and
Hypothesis Testing
Presented by:
Ariate, Hanny Grace A.
Hayagan, Jeneva T.
Paraon, Aileen Mae R.
Sampling Distributions
The probability distribution of a statistic.
z=
If a large number of samples of a given size are
selected from a normally distributed population, or if a
large number of samples of a given size is greater than or
equal to 30 are selected from a population which is not
normally distributed, and the sample means are
computed. Their percentages indicate the areas of the
regions.
34.13% 34.13%
13.59% 13.59%
2.28% 2.28%
It’s important to remember two things when you use
central limit theorem:
Thus,
Example:
Suppose salaries at a very large corporation has a mean
of 62,000 pesos and a standard deviation of 32,000
pesos. If 100 employees are randomly selected, what is
the probability their average salary exceeds 66,000
pesos?
Solution:
Thus,
Hypothesis Testing
Concepts of Hypothesis Testing
There are two hypotheses-the null and alternative hypothesis. The
alternative hypothesis only comes into play when the null
hypothesis is rejected.
We begin with the unusual thing: that the claim is correct or the
null hypothesis is true.
Gather enough evidence to disprove the null hypothesis.
There are two possible decisions: to reject the null hypothesis or to
accept the null hypothesis.
Two errors may also occur. Rejecting a true null hypothesis or
failing to reject a false null hypothesis.
Five-Step Procedure for Testing of
Hypothesis
Hypothesis - is a statement about the population
developed for the purpose of testing (Keller and
Warrack, 2004).
a. State the null hypothesis and alternate hypothesis.
The null hypothesis is denoted by . The subscription O
in H gives us a guide on how to state the null hypothesis. It
implies “no difference”. The null hypothesis should be
stated by saying “There is no significant difference…” or
“There is no relationship…” or “There is no change…”
Rejection of the null hypothesis directly results
acceptance of the alternate hypothesis denoted by The
alternate hypothesis only comes into play once the
decision calls for the rejection of null hypothesis.
To state alternate hypothesis, we negate the null
hypothesis. Example, “There is a significant difference …”
The alternate hypothesis will tell us if the test is one-
tailed test or two-tailed test.
One-tailed test is a test where the area of rejection is at
either side. Two-tailed test on the other hand is a test
where the areas of rejection are both of a distribution.
One-tailed Test
Area of
rejection
Area of
rejection
Two-tailed Test
Area of Area of
rejection rejection
Example
A test was administered to two groups of students- the
HRM students group and the tourism group. At the
significance level, is there difference between the scores
obtained by the two groups of students?
Solution:
There is no significant difference between the
scores obtained by the two groups of students.
There is significant difference between the
score obtained by the two groups of students.
If we say, “there is difference…”, this is susceptible of two
meanings: (a) that the mean score of the HRM student
group may be higher then the mean score of tourism
student group; and (b) that the mean score of HRM
student group may be lower than the mean score of
tourism student group.
If the alternate hypothesis poses two meanings, this is
called non-directional. In this case, we use the two-
tailed test.
Example
A teacher employed two different teaching strategies in presenting
her lesson: lecture and discussion method. After the presentation, a
30-point quiz was given. The mean score of the students where the
discussion method was the strategy used was found out to be 25 with a
standard deviation of 3. the mean score of the students where the
lecture method was used found out to be 19 with standard deviation
of 3.2. at the .01 significance level, can we conclude that the discussion
method is more effective than the lecture method?
Solution:
There is no significant difference between the two teaching
strategies. Or
The discussion method is as effective as the lecture method.
Discussion method is more effective than the lecture
method.
Notice that the alternate hypothesis is very specific. The
problem only requires to the test that the discussion
method is more effective than the lecture method. The
alternate hypothesis is directional. In this case we use the
one-tailed test.
b) Set the desired level of significance
The level of significance is also termed as alpha or level of
risk and is denoted by . The alpha level can be a value less
than or equal to 10%. Traditionally, what is being used
0.05 or 5% for consumer research project, .01 for quality
assurance and .10 for political polling(Keller and
Warrack, 2004). But you as the researcher can choose the
level of significance you want to use in your study.
Level of risk is the probability of rejecting a true null
hypothesis.
If the null hypothesis is true and is rejected, it is called
type I error. And if the null hypothesis is false and is
accepted, it is called type II error.
Decision
Null Hypothesis
Reject Accept
Area of
Acceptance
Area of Area of
rejection rejection
.025 .475 .475 .025
1. Type of Data
Data can be classified as quantitative or qualitative. Nominal
and ordinal levels fall under qualitative. Interval and ratio fall
under quantitative. If we are dealing with quantitative data, we
will used the parametric tests. These are the t-test, z-test or F-
test. If we are dealing with qualitative data, we will use non-
parametric tests. To mention some, we have Chi-square, Mann-
Whitney U Test, etc.
2. Purpose of the study
We may want to conduct a study to determine whether
relationship exists between two variables.
3. Number of Observations
If population standard deviation is not given, we may use t-
test or z-test depending upon the number of observations.