Professional Documents
Culture Documents
Module 4
II. Overview
In this module, we will explore some statistical processes that will help us figure
out whether the assumptions of research can be held correct or not using the data
from the sample understudy. Upon completion of this module, you will be able to make
your testable assumptions about the variable(s) in question and conduct hypothesis
testing.
III. Objectives
IV. Discussion
Hypothesis Testing
Before embarking on the entire process, there are some terminologies and concepts
you need to get familiarized with. The following are the most common ones:
2
PSYCH 2104 (PSYCHOLOGICAL STATISTICS)
In the actual research process, the null hypothesis is something that may or may
not true. You can reject or accept it depending on the result of the analysis. However,
there are two possible errors or faulty decisions a researcher can commit: Type 1 and
Type 2 errors. Type 1 error occurs when a researcher rejects the null hypothesis while in
fact, it is true. In other words, you commit a type 1 error when you say that there is a
significant difference between the parameters while there is none. In contrast, you
commit the type 2 error if you fail to reject the null hypothesis when it is false. In other
words, Type 2 error occurs when you say that there is no significant difference between
the parameters while in fact, there is.
The critical value indicates a certain point in the distribution that separates the
critical region. The critical region aids the researcher to determine whether a significant
difference exists between the parameters and that the null hypothesis should be rejected.
The critical region can be found on either side of the distribution. This depends on
whether the test is one-tailed or two-tailed. For instance, in a left-tailed test, an alpha of
.01, the critical value is -2.33, and the area from that point toward the extreme left side
will be the critical region. In a right-tailed test, the alpha level of .01 has a critical value
of 2.33. Notice that the only difference is the sign. Therefore, all left-tailed tests have
negative critical values and all the right-tailed tests have positive critical values. In the
two-tailed test, on the other hand, the critical value or the critical region is located on
either side of the distribution.
3
PSYCH 2104 (PSYCHOLOGICAL STATISTICS)
In a two-tailed test, the rejection region is located on either side of the distribution.
The direction is not specified. This is the reason why the two-tailed test is also known as
the non-directional hypothesis. The symbol will look like this Ho: μ1 = μ2; H1: μ1 ≠ μ2.
This happens in studies where a researcher predicts that there is a significant effect but
he/she does not know what will be the effect is (positive or negative). Another thing to
4
PSYCH 2104 (PSYCHOLOGICAL STATISTICS)
remember is that, in two-tailed tests, the alpha level or the significance level is divided
into 2. So if the alpha level is .05, the rejection region will be 0.025 on both sides of the
distribution which has the z score of -1.96 on the left and +1.96 on the right. These
values are the critical values from which the rejection of the null hypothesis will be based
upon. Let us see how one and two-tailed tests look in the examples below.
The common statistical test being conducted in a large sample size is the Z test.
This is applicable if the population is equal to or greater than 30. The following is the
formula you can use (Z Test: Formula & Example, 2017).
Z = x̅ - μ/ σ/√n
Where:
Example 1
Now, let’s work on one sample problem using a one-tailed test (left-tailed test).
Suppose that you heard a claim that students spend a lesser amount of time
watching television than the general public. The existing claim also states that on
average, the general public spends 28 hours per week on their TV set with a standard
deviation of 3. To your curiosity, you conducted a study wherein 40 college students
participated. From the data, you found out that the average hours spend by the
participants watching television is 24 hours. Using an alpha level of .01, test the validity
of the claim.
Applying the steps for conducting hypothesis testing, you can follow the steps below:
5
PSYCH 2104 (PSYCHOLOGICAL STATISTICS)
H0: μ ≥ 28
H1: µ < 28 (less than the existing claim)
Or
H0: μ1 ≥ μ2
H1: µ1 < μ2
Using the alpha of .01, and since this is a left-tailed test, the critical region is at
the left side of the distribution equivalent to Z = -2.33 (this value is derived from the z
distribution table).
Note: To find the z score of any given alpha level, look for the value nearest to
the alpha, in this case, .01 on the Area Beyond Z, and then look for its corresponding
value on the Z column. In this case, the closest value to .01 is .0099 and the
corresponding Z value is 2.33. Since the test is left-tailed, it should be negative (-2.33).
Also, note that there are no negative values in the table because the other half of the
distribution has equal value or area.
6
PSYCH 2104 (PSYCHOLOGICAL STATISTICS)
As you can see, the test value is -8.43 less than the critical value of -2.33 and falls
in the rejection region, therefore, the null hypothesis should be rejected. (Note: If the
test value is greater than -2.33, do not reject the null hypothesis)
The result provides evidence to support the existing claim that college students
spend less amount of time watching television than the general public.
Example 2
Now let’s use the same example using a one-tailed test (right-tailed test). Suppose
that you heard a claim that students spend more time watching television than
the general public. The existing claim also states that on average, the general public
spends 24 hours per week on their TV set with a standard deviation of 3. To your curiosity,
you conducted a study wherein 40 college students participated. From the data, you
found out that the average hours spend by the participants watching television is 28
hours. Using an alpha level of .01, test the validity of the claim.
H0: μ ≤ 24
H1: µ > 24
Or
H0: μ1 ≤ μ2
H1: µ1 > μ2
7
PSYCH 2104 (PSYCHOLOGICAL STATISTICS)
Using the alpha of .01, and since this is a right-tailed test, the critical region is at
the right side of the distribution equivalent to Z = 2.33 (this value is derived from the z
distribution table).
Note: To find the Z score of any given alpha level, look for the value nearest to
the alpha, in this case, .01 on the Area Beyond Z in the table and then look for its
corresponding value on the Z column. In this case, the closest value to .01 is .0099 and
the corresponding Z value is 2.33.
As you can see, the test value is 8.43 greater than the critical value of 2.33 and
falls in the rejection region, therefore, the null hypothesis should be rejected. (Note: If
the test value is less than 2.33, do not reject the null hypothesis )
8
PSYCH 2104 (PSYCHOLOGICAL STATISTICS)
The result provides evidence to support the existing claim that college students
spend more time watching television than the general public.
Example 3
Now let’s use the same example in a two-tailed test. Suppose that you heard a
claim that the time spent on watching TV greatly differs between students and
the general public. The existing claim also states that on average, the general public
spends 24 hours per week on their TV set with a standard deviation of 3. To your curiosity,
you conducted a study wherein 40 college students participated. From the data, you
found out that the average hours spend by the participants watching television is 28
hours. Using an alpha level of .01, test the validity of the claim.
H0: μ = 24
H1: µ ≠ 24
Or
H0: μ1 = μ2
H1: µ1 ≠ μ2
Using the alpha of .01, and since this is a two-tailed test, the alpha will be divided
into 2 which is equal to .005. The critical region is below Z = -2.58 and above or greater
than Z = 2.58 (this value is derived from the z distribution table).
Note: To find the z score of any given alpha level, look for the value nearest to
the alpha, in this case, .005 on the Area Beyond Z in the table and then look for its
corresponding value on the Z column. In this case, the closest value to .005 is .0049 and
the corresponding Z value is 2.58.
9
PSYCH 2104 (PSYCHOLOGICAL STATISTICS)
As you can see, the test value is 8.43 greater than the critical value of 2.58 and
falls in the right rejection region, therefore, the null hypothesis should be rejected. (Note:
If the test value is less than 2.58, do not reject the null hypothesis )
The result provides evidence to support the existing claim that the time spent in
watching TV differs between college students and the general public.
Video Presentations
Hypothesis Testing
Intro to Hypothesis Testing in Statistics
Intro to Hypothesis Testing in Filipino
Journal Articles
Hypothesis Testing
10
PSYCH 2104 (PSYCHOLOGICAL STATISTICS)
Online Resources
Hypothesis Testing
What is Hypothesis Testing
V. Assessment
Module Quiz 4
Read the following statement carefully and try to answer the given question.
Follow the exact format from the previous examples. This is a 30-point quiz.
Rubrics
The new memory enhancer X recently gained a reputation in the market as the
most effective memory enhancer ever produced by XY Pharma. The recorded average
memory level among people who had taken the enhancer was 95 with a standard
deviation of 10. As a neuroscientist, you want to test the claim and conducted a study
and measure the memory level of 50 participants. The average memory level of the
participants came out to be 80. Is there enough evidence to support the claim about the
new memory enhancer?
VI. References
Aron, A., Coups, E. J., & Aron, A. N. (2013). Statistics for Psychology: Sixth Edition.
Pearson Education, Inc.
11