TH2301

DATA ANALYSIS AND RESULTS DISCUSSION IN RESEARCH

Hypothesis Testing
Hypothesis Testing is a procedure on sample evidence and probability used to test claims regarding one or
more populations’s characteristics.
Procedures for Testing Hypothesis
1. Identify the terms or phrases used in Research Questions that can trigger hypothesis testing.
• In research, different terms or phrases in research questions prompt hypothesis testing.
These terms suggest relationships or comparisons that researchers explore and verify using
statistical methods. Here are some examples:
o Correlation: Research questions investigating the degree and direction of a
relationship between variables. The usual term used in this context is a significant
relationship.

Example: Is there a significant relationship between hotel ratings and customer

satisfaction levels?

o Difference: Research questions that identify differences between groups or

conditions. The usual term used in this context is a significant difference.

Example: Is there a significant difference in dining preferences among international

tourists based on cultural backgrounds?

2. State the null(𝐻0 ) and alternative (𝐻𝑎 ) hypothesis.

• In conducting a research study, if a hypothesis test is used to support a claim, the claim must
be stated to become the alternative hypothesis, so it cannot contain the condition of
equality.
• The research designs below usually have hypothesis testing. One (or some) of the Research
Questions under these designs trigger a hypothesis test.
o Quantitative Comparative
o Quantitative Experimental
o Mixed Methods (as applicable)
Sample:
Research Title: Assessing The Food and Service Quality of Typica Coffee in Taytay, Rizal
Research Question: Is there a significant difference between the food & service quality and
product loyalty among customers of Typica Coffee?
𝐻0 : There is no significant difference between the food & service quality and product loyalty among
customers of Typica Coffee.
𝐻𝑎 : There is a significant difference between the food & service quality and product loyalty among
customers of Typica Coffee.

01 Handout 2 *Property of STI

Page 1 of 6
 TH2301

Statistical Treatment
The statistical tests allow researchers to make inferences because they can show whether an observed
pattern is due to intervention or chance. There is a wide range of statistical tests. The decision of which
statistical test to use depends on the research design, the distribution of the data, and the variable type.
Generally, if the data is normally distributed, parametric tests should be used. If the data is non-normal, non-
parametric tests should be used. The table below shows a list of just a few common statistical tests and their
uses.

Type of Test Use Sample Scenarios

Correlational: These tests look for an association (or relationship) between two variables.
Pearson Tests for the strength of the association Assessing if there is a relationship between
Correlation between two continuous variables. hotel occupancy rates and revenue
Analysis of Variance: These tests look for the difference between the means of variables
Dependent Tests for the difference between two Evaluating the sales performance before
t-Test variables from the same population and after a marketing campaign
Independent Tests for the difference between the same Comparing the customer satisfaction
t-Test variable from different populations between two (2) different branches of a
restaurant
ANOVA Tests for the difference between group Analyzing the performance of employees
means after any other variance in the working in the different departments of a
outcome variable is accounted for hotel (e.g., front desk, housekeeping, and
food and beverage)

P-Value Analysis
Using the p-value approach
Decision rule: Reject the null hypothesis if the computed p-value is less than or equal to the set significance
level; otherwise, do not reject the null hypothesis.

Example:
If the level of significance 𝛼=0.05,
P-value Decision
0.01 Reject 𝐻0
0.05 Reject 𝐻0
0.10 Failed to reject 𝐻0

Sample statistical results with p-values:

Table 1.
Significant Difference Between Food & Service Quality and
Product Loyalty among Customers of Typica Coffee
Mean SD t Sig. H0 VI
Food Quality 3.38 0.12
1.266 0.253 FR NS
Product Loyalty 3.31 0.08
Environment Quality 3.33 0.10
0.429 0.681 FR NS
Product Loyalty 3.31 0.08
Interaction Quality 3.53 0.07 5.500 0.000 R S

01 Handout 2 *Property of STI

Page 2 of 6
 TH2301

Product Loyalty 3.31 0.08

Satisfaction 3.35 0.03
1.335 0.212 FR NS
Product Loyalty 3.31 0.08
Self-Efficacy 3.34 0.06
1.012 0.335 FR NS
Product Loyalty 3.31 0.08
Legend:
t – t-value Sig. – level of significance H0 – null hypothesis VI – verbal interpretation
R – reject S – significant FR – failed to reject NS – not significant

The ‘Sig.’ column above shows the obtained p-values in each pair of variables tested. It can be seen from the
table that overall, since most of the variables obtained a p-value that is more than 0.05 level of significance,
the null hypothesis failed to be rejected. This means that there is no significant difference between food &
service quality and Typica customer’s product loyalty with respect to the stated variables. However, there is
a significant difference between interaction quality and product loyalty of its customers.

Presentation of Data
The effective presentation of data is essential for clearly and efficiently communicating research findings.
This elevates the overall quality of the research and enhances its impact on the target audience since it
provides reliable evidence to support the study. Below are some tables used to present research data.

Frequency and Percentage Distribution Tables

A Frequency and Percentage Distribution Table is shown to illustrate the demographic profile of the
respondents, depending on what they have identified. Some of which include age, sex, gender, monthly
family income and more. The table below shows how these are presented.

Table 2.
Frequency and Percentage Distribution on the Profile of the Respondents in Terms of Gender
Gender f p Rank
Female 72 52.55% 1
Male 65 47.45% 2
Total 137
Legend:
f – frequency
p – percentage

The table shows that most of the customers are female, with a frequency of 72 and a percentage of 52.55%,
while the male customers have a frequency of 65 and a percentage of 47.45%.

01 Handout 2 *Property of STI

Page 3 of 6
 TH2301

Computed Mean Tables

A table showing the computed mean usually illustrates the general view of all respondents in a research,
especially when a Likert Scale is used.

Table 3
Computed Mean on the Assessment Level of the Respondents on the
Food and Service Quality of Typica Coffee
Food Quality Mean VI Rank
1. Extensive menu 3.45 E 2
2. Presentation of food and beverage 3.55 E 1
3. Overall taste of food and beverage 3.36 E 3
4. Nutritional value of food and beverage 3.31 E 4
5. Availability of food and beverage 3.24 G 5
Average 3.38 E
Legend:
1.00 to 1.74 – Needs Improvement (NI);
1.75 to 2.49 – Fair (F);
2.50 to 3.24 – Good (G); and
3.25 to 4.00 – Excellent (E)

The table shows that among the respondents, they value the presentation of food and beverage with a mean
of 3.55 and a verbal interpretation of excellent since it is ranked the highest while the availability of food and
beverage ranked the lowest with a mean of 3.24 and a verbal interpretation of good.

It implies that customers select this coffee shop because of its unique presentation of their food and
beverages rather than having a wide array of food and beverage.

This supports the study of (author, year) that visually appealing food is one of the primary reasons why
customers patronize a food establishment.

Composite Tables
A composite table typically refers to a table that integrates or combines data from multiple variables or
dimensions into a unified format for analysis or presentation.

Table 4.
Composite Table on the Assessment Level of the Respondents on the
Food & Service Quality of Typica Coffee
Mean VI Rank
1. Food Quality 3.38 E 2
2. Environment Quality 3.33 E 5
3. Interaction Quality 3.53 E 1
4. Satisfaction 3.35 E 3
5. Self-Efficacy 3.34 E 4
Average 3.39 E

01 Handout 2 *Property of STI

Page 4 of 6
 TH2301

Legend:
1.00 to 1.74 – Needs Improvement (NI);
1.75 to 2.49 – Fair (F);
2.50 to 3.24 – Good (G); and
3.25 to 4.00 – Excellent (E)

The table shows that among the respondents…

It implies that… why is that the result? What could have happened?

This supports the study of…

Significant Difference Tables

A table that shows the significant difference between two (or more) variables reveals the data that
concludes a hypothesis test.

Table 5.
Significant Difference on the Assessment of the Respondents on the
Service Quality of Kampo de Angono Restaurant Using SERVQUAL
Dimensions in Terms of Age

Age Mean SD F Sig. H0 VI

11 to 26 years old 3.31 0.09
Assurance 27 to 42 years old 3.18 0.10 2.452 0.128 FR NS
43 to 58 years old 3.30 0.12
11 to 26 years old 3.12 0.10
Empathy 27 to 42 years old 3.04 0.07 5.256 0.023 R S
43 to 58 years old 3.23 0.11
11 to 26 years old 3.27 0.09
Reliability 27 to 42 years old 3.07 0.15 3.035 0.086 FR NS
43 to 58 years old 3.08 0.19
11 to 26 years old 3.22 0.21
Responsiveness 27 to 42 years old 3.03 0.21 1.204 0.334 FR NS
43 to 58 years old 3.12 0.19
11 to 26 years old 3.34 0.13
Tangibles 27 to 42 years old 3.41 0.12 1.400 0.284 FR NS
43 to 58 years old 3.46 0.09
11 to 26 years old 3.25 0.14
Overall 27 to 42 years old 3.15 0.19 2.728 0.072 FR NS
43 to 58 years old 3.24 0.19
Legend: F – F-value Sig. – level of significance H0 – null hypothesis
VI – verbal interpretation R – reject S – significant
FR – failed to reject NS – not significant

01 Handout 2 *Property of STI

Page 5 of 6
 TH2301

It can be seen from the table that overall, since the p-value is greater than 0.05 level of significance, the null
hypothesis failed to be rejected. This means that there is no significant difference in the assessment of the
respondents on the service quality of Kampo De Angono restaurant using SERVQUAL dimensions in terms of
age.

It implies that… why is that the result? What could have happened?

This supports the study of…

References:
Bevans, R. (2020). Hypothesis testing. Retrieved from: https://cyfar.org/types-statistical-tests
Frost, J. (N.D.) Measures of variability. Retrieved from https://statisticsbyjim.com/basics/variability-range-interquartile-
variance-standard-deviation
Gaur, A., Gaur, S. (2009). Statistical method for practices and research. Singapore: SAGE Publication Inc.
Malvar R. (2019). Statistics applied to research. Retrieved from: STATISTICS-LECTURE-PART-4.pdf
Pescasio, A., et al (2017). Math ideas and life application. Quezon City: ABIVA Publishing House, Inc.
Philippine Statistical Authority (2020). 2017 Annual survey of Philippine business and industry- Economy wide all
establishments: Final result. Retrieved from: https://psa.gov.ph/content/2017-annual-survey-philippine-
business-and-industry-aspbi-economy-wide-all-establishments
Punzalan, J. (2018). Statistics and Probability. Malaysia: Oxford Publishing.
Siegel D. (2002). Educational research basics. Retrieved from: https://researchbasics.education.uconn.edu/t-test/#
Sullivan, M. (2017). Informed decision using data: Fifth edition: Pearson Education

01 Handout 2 *Property of STI

Page 6 of 6