College of Communication

Polytechnic University of the Philippines

Anonas, Sta. Mesa, Maynila, Kalakhang Maynila

Statistical Analysis with Software Application

(STAT 20053) SCORE/ GRADE:

FINAL EXAMINATION

Name: Pauline Joy E. Ilig. Section: BSMA 2-3 Schedule: Wed. 2:30-7:30PM Date: Feb. 15, 2023

Test 1: Explain the following questions.

1. Differentiate the dependent and independent samples?

Dependent or paired samples examine a single set of variables in a single set of cases.
Two groups of instances are used by independent samples to compare various sets of variables.
Whether a sample is dependent or independent must be understood. They are both employed
to measure a variety of items and provide information on hypothesis testing. Scientific studies
frequently compare two hypotheses: a null hypothesis (which typically asserts that there is no
effect) and an alternative hypothesis (which asserts that the effect of interest exists) (Anderson,
2019).

The data present two situations for dependent samples and are related to one another.
The first scenario compares results from before and after. In the second scenario, one sample
is the main focus, and an observation is matched with an observation from a different sample.
“We are able to limit the variation in the sampling distribution by using dependent samples” (Lind,
Marchal, & Wathen, 2015). For practical testing, values must have comparable components.
The mean of the distribution of differences in each value is used for hypothesis testing in
dependent samples.

To determine the mean of two samples, independent samples gather data on two
variables. For reliable findings, it is assumed that the population standard deviations are equal.
Except for situations when samples are larger than 30, populations typically have unknown
means and standard deviations. The difference between the sample means divided by the
standard error is a t-score test statistic. Testing your hypotheses can help you comprehend
 samples and potential relationships better. Remember that mistakes can happen when testing
something.

2. Discuss how p-values are used in the test of hypothesis.

The probability that you would have discovered a specific collection of observations if the
null hypothesis were true is expressed as a number called the p value, which is determined by
a statistical test. In order to determine whether to reject the null hypothesis, P values are utilized
in hypothesis testing. The smaller the p value, the more inclined you are to reject the null
hypothesis.

The probability value, or p value, indicates the likelihood that your data could have been
true under the null hypothesis. This is accomplished by estimating the probability of your test
statistic, which is the figure obtained from a statistical analysis of your data. If the null hypothesis
of your statistical test were true, the p value would inform you how frequently you would
anticipate seeing a test statistic that was as extreme or more extreme than the one determined
by your test. As the test statistic derived from your data deviates more from the distribution of
test statistics anticipated by the null hypothesis, the p value decreases. A test statistic that is at
least as dramatic as the one you discovered would appear 5% of the time if the null hypothesis
were true, according to the definition of the p value, which is a proportion.

The P value indicates the likelihood that randomly placed dots will produce an association
with a similar strength. We are more confident in forecasting the line when the p-value is lower.
It is decided before the test’s execution if it has anything to do with the alpha level’s relevance.
The null hypothesis is rejected if the p-value is determined to be smaller than the alpha test,
which is regarded as statistically significant, i.e., startling or odd.
Test 2: For the following situations, identify the appropriate Null and Alternative Hypothesis.
1. A company which manufactures one-half inch bolts selects a random sample of bolts to determine
if the diameter of the bolts to determine if the diameter of the bolts differs significantly from the
required one-half inch.

Null: The diameter of the bolts does not differ significantly from the required one-half inch.
Alternative: The diameter of the bolts differs significantly from the required one-half inch.

2. A company which manufactures safety flares randomly selects 100 flares to determine if the flares
last at least three hours on average.

Null: The flares do not last at least three hours on average.

Alternative: The flares last at least three hours on average.

3. A consumer group believes that a new sports coupe gets significantly less miles to the gallon than
advertised on the sales stickers. To confirm this belief, they randomly select several of the new
coupes and measures the miles per gallon.

Null: The new sports coupe does not get significantly less miles to the gallon than advertised on the
sales stickers.
Alternative: The new sports coupe does get significantly less miles to the gallon than advertised on
the sales stickers.

4. A spokesperson for a popular diet claim that the average weight lost for someone on the diet will
be at least 15pounds over a two-month period. The amount of weight lost for each person in a sample
of 10 people on the diet is determined in order to try to refute the claim of the diet spokesperson.

Null: The average weight lost for someone on the diet is not at least 15 pounds over a two-month
period.
Alternative: The average weight lost for someone on the diet is at least 15 pounds over a two-month
period.

5. Random and independent samples of younger (age≤30) and older (age>30) automobile drivers are
chosen and asked whether they have had a speeding ticket in the past 12 months. It is intended to
show that younger drivers are more likely than older drivers to have had a speeding ticket in the past
12 months.

Null: There is no age restriction in order to receive a speeding ticket.

Alternative: Younger drivers are likely to have a speeding ticket than the older drivers.

6. Do women have a shorter reaction time than men when exposed to a certain stimulus? Random
and independent samples of 10 men and 10 women are included in an experiment that measures
reaction time to the stimulus.

Null: The is no gender specification regarding those that has a shorter reaction time when exposed to
a certain stimulus.
Alternative: Women have no shorter reaction time than men when exposed to a certain stimulus.

7. A tire company test 68 of its new premium tires to determine if the average life length of the tire is
more than average life length of its major competitor’s best tire. The average life length of the
competitor’s tire is 63,000 miles.

Null: The average life length of the tire is not more than average life length of its major competitor’s
best tire.
Alternative: The average life length of the tire is more than average life length of its major
competitor’s best tire.

8. An elementary statistics student conducts an experiment in order to show that a coin from a magic
kit is biased. The student flips the coin 500 times.

Null: The coin from a magic kit is not biased.

Alternative: The coin from a magic kit is biased.

Test 3: Computation. Show necessary solutions.

1. A consulting firm implemented a quality program in an effort to improve the productivity level of its
employees. All employees received two and one-half days of intensive training on the concepts of
total quality management, with an emphasis on process mapping. In addition to the initial training
each employee participated on a team (a QET-Quality Enhancement Team or CIT Continuous
Improvement Team) which focused on either breakthrough improvements in a process-QET or
continuous improvement of a process CIT. Each team was given the responsibility and the authority
 to implement the improvements in the process which it deemed necessary. A year after the initial
training, the management of the company conducted a study to see if the productivity level of its
employee had improved. The productivity level of employees is randomly selected and the realized
billable hours are calculated before the initial quality training and one year after the initial quality
training. The results of the survey are shown below.

a) What level of measurement does the data possess?

• This data has ratio and interval levels of measurement. Since the ratio between the
number of realized billable hours and total output and the degree of productivity of the
employees before and after the program of rigorous quality training are being measured.

b) Determine whether the productivity level of the employees increase after the intensive
training. Use an alpha of 0.05. Perform the six steps of hypothesis testing

Step 1: State the Null and Alternative Hypothesis

• Null Hypothesis: 𝐻𝒐 : 𝜇1 ≥ 𝜇2
The program of intensive quality training will not help to improve the productivity level of the
employees in basis of realized billable hours.
• Alternative Hypothesis: 𝐻𝑎 : 𝜇1 < 𝜇2
The program of intensive quality training will help to improve the productivity level of the employees in
the basis of realized billable hours.
Step 2: Set the Level of Significance or Alpha Level (α)
- Alpha level of 0.05

Step 3: Determine the Test Distribution to Use.

Dependent Variable: Productivity level of Employees in the basis of Realized Billable Hours
Independent Variable: Treatment (Before and After)
Since we are comparing the means of two related groups, we will use the dependent sample t-test.

Step 4: Calculate Test Statistic or p - value. (I used Excel or Other Software)

• p-value Computation performed for Realized Billable Hours BEFORE the Quality
Training data.
• p-value Computation performed for Realized Billable Hours AFTER the Quality Training
data.

• p-value Computation performed for DIFFERENCE

 • p-value Computation performed for ALL the DATA (FINAL ANSWER)

Step 5: Make Statistical Decision

• p-value Result for Realized Billable Hours BEFORE the Quality Training data.
• p-value Result for Realized Billable Hours AFTER the Quality Training data.

• p-value Result for DIFFERENCE

• p-value Result for ALL the DATA (FINAL ANSWER)

 t-Test: Paired Two Sample for Means

Variable 1 Variable 2
Mean 25.94 25.26
Variance 28.17333333 34.7525
Observations 25 25
Pearson Correlation 0.136343412
Hypothesized Mean Difference 0
df 24
t Stat 0.461005472
P(T<=t) one-tail 0.324472612 Failed to Reject Null Hypothesis
t Critical one-tail 1.71088208
P(T<=t) two-tail 0.648945225
t Critical two-tail 2.063898562

Step 6 : Draw Conclusion

There is no enough evidence that the intensive quality training improves the level of productivity of
the employees.
Proper Presentation of Result:
Indicator Treatment Mean t-value p-value Decision Remarks
Productivity BEFORE 25.94
level of
Employees in Failed to Not
-0.461005 0.32447
the basis of AFTER 25.26 reject 𝐻𝑜 Significant
Realized
Billable Hours

2. A personnel director is interested in studying the relationship (if any) between age and salary.
Sixteen employees are randomly selected and their age and salary are recorded.
a) What level of measurement does the data possess?
• Since this data measures continuous variables like age and salary, ratio and interval levels of
measurement are included.
b) Determine if there is a significant relationship between age and salary. Use an
alpha of 0.05. Perform the six steps of hypothesis testing.
 Age Salary Age Salary
25 22000 49 39000
55 45000 37 45000
27 43000 62 60000
30 30000 40 35000
22 24000 35 34000
33 53000 29 30000
19 18000 58 73000
45 38000 52 42000

Step 1: State the Null and Alternative Hypothesis

• Null Hypothesis: There is no significant relationship between the age and salary.

• Alternative Hypothesis: There is a significant relationship between the age and salary.

Step 2: Set the Level of Significance or Alpha Level (α)

- Alpha level of 0.05

Step 3: Determine the Test Distribution to Use.

Dependent Variable: Age
Independent Variable: Salary
Since we are testing the significant relationship of two variables, we will use Pearson r.

Step 4: Calculate Test Statistic or p - value. (Using Excel or Other Software)

Step 5: Make Statistical Decision

Step 6: Draw Conclusion

There is enough evidence to conclude that there is a significant relationship between the age and salary
of the sixteen employees.
Proper Presentation of Result:
AGE
Pearson r Description P - value Decision Remarks
Strong and
SALARY
0.7707 Direct .000475 Reject 𝐻𝑜 Significant
Correlation