Inferential Statistics:

[Measures of Differences, Measures of Relationship,

and Measures of Effects]
(Formulation of Test Hypotheses and Decision Making)

BENJAMIN L. AMBROSE
Instructor I
Part I: What is Hypothesis?
❑A hypothesis is a proposed explanation/prediction
(difference) for a phenomenon/event, relationship
or effect/impact between variables. It serves as a
foundation for scientific research and guides
experiments to test its validity(Kerlinger, 1986).
❑). A well-formulated hypothesis can be tested
through observation or experimentation.
 Role of a Hypothesis in Research
❑It gives the study a clear purpose and helps
identify what data to collect – focus and direction.

❑It allows researchers to test and either accept or

reject a proposed idea – basis for experiments.

❑It provides a reference point to compare

experimental outcomes –framework for analysis.
 Test of a Hypotheses
1. Non-directional Hypothesis (Null (H₀) & Alternative (Hₐ) Hypotheses
❑A statement that suggests is or no difference, relationship, or effect exists.
❑It does not tell the direction of relationship between or among variables (i.e., direct
or inverse).
Example:
1. There is no significant difference in academic performance between students who study in the morning and
those who study at night.
2. There is a relationship between internet use and time spent with the family.
3. Students who study in the morning perform better academically than those who study at night.
2. Directional Hypothesis
❑It states very explicitly the direction of relationship whether direct and inverse.
Example:
1. The more time spent to internet, the lesser is the time spent for the family.
2. Students enrolled in Public Administration courses with a practical internship component will
demonstrate higher levels of leadership skills compared to those without an internship component.
3. Firms that adopt cloud-based accounting software will show a higher rate of accuracy in financial
reporting compared to firms using traditional accounting methods.
Turning Research Questions into Testable Hypothesis
Accountancy and Accounting Research Problem
Title of Research :Financial Literacy and Financial Management Practices of
Selected Students Boarders in Sorsogon State University BC
Proposition :Financial Literacy is affected by Financial Management
Practices
Research Questions (Actual):
SOP 3. What is the correlation between financial literacy
and financial management of the student boarders of
Sorsogon State University – Bulan Campus?
Hypothesis 1 :There is a correlation between financial literacy and financial
management of the student boarders of SorSU BC.
Hypothesis 2 :There is a correlation between financial skills and student’s
personal finance practices.
Hypothesis 3 :As financial skills of the students increase, the student’s personal
finance practices decreases.
Turning Research Questions into Testable Hypothesis
Public
Title of Research
Administration Research Problem
:Political Awareness and Political Involvement of the Students
at Sorsogon State University Bulan Campus
Proposition :Political awareness is affected by their political involvement
Research Questions (Actual):
SOP 4. Is there significant correlation between the level of political
awareness and political involvement of the students at Sorsogon
State University – Bulan Campus students?
Hypothesis 1 :There is correlation between level of political awareness and
political involvement of the student of SorSU BC.
Hypothesis 2 :There is correlation between student’s informed decisions and
student’s engagement.
Hypothesis 3 :As student’s informed decisions increase, the student’s engagement
decreases.
 Characteristics of a Good Hypothesis
❑A strong hypothesis should:
[Link] Testable: It must be possible to confirm or refute it through
experiments or observations.
[Link] Measurable: The variables involved should be quantifiable or
observable.
[Link] Clear and Specific: It should clearly define the relationship or effect
being tested.
[Link] Based on Research: Use prior knowledge, existing theories, or
preliminary data as a foundation.
 Common Mistakes to Avoid
❑Common Mistakes to Avoid
1. Being Too Broad: Avoid vague statements.
Example : Exercise is good for health.
Instead : Regular aerobic exercise reduces blood pressure.
Example 2 : Training programs make employees better at their jobs.
Instead : Public officials who attend a leadership training program will improve their
decision-making efficiency compared to those who do not.
2. Making It Unmeasurable: Avoid hypotheses that cannot be tested or observed.
Example 1 : If students are motivated, then they will succeed.
Revised : Students who participate in motivational workshops will show a higher GPA
improvement compared to those who do not.

Example 2 : Firms with better management practices are more successful.

Revised : Firms that implement quarterly financial performance reviews will achieve
higher profit margins than those that only conduct annual reviews.

3. Ignoring the Variables: Clearly define what you are testing and measuring.
 Part II: Hypothesis Errors in Decision-Making
❑In hypothesis testing, researchers aim to
determine whether to reject or fail to
reject the Null Hypothesis (H₀).
However, the decision-making process is
not foolproof and may lead to errors.

❑These errors, known as Type I and Type II

errors, occur due to the inherent
uncertainty in statistical inference.
1. Type I Error (False Positive)
❑A Type I error occurs when the null hypothesis (H₀) is rejected even though
it is true.
Example:
Conclusion: Concluding that a new teaching method improves student
performance when it actually does not.
Implication: This type of error leads to a false alarm, causing
researchers to believe there is an effect or relationship when none
exists.

❑Probability of Occurrence (α):

The significance level (commonly set at 0.05) determines the
likelihood of a Type I error. A lower α reduces the probability of
making this error but may increase the likelihood of a Type II error.
2. Type II Error (False Negative)
❑A Type II error occurs when the null hypothesis (H₀) is not rejected even
though it is false.
Example:
Conclusion: Concluding that a training program has no
impact on employee productivity when it actually does.
Implication: This type of error leads to a missed
opportunity, causing researchers to overlook a genuine
effect or relationship.

❑Probability of Occurrence (β): The power of the test (1 - β) affects the

likelihood of avoiding a Type II error. Increasing sample size or effect size
can reduce β.
 Factors Influencing Hypothesis Errors
1. Sample Size. A larger sample size reduces variability and increases the
reliability of the results, decreasing the likelihood of Type II errors.

2. Significance Level (α). Setting a lower α (e.g., 0.01 instead of 0.05)

reduces the chance of a Type I error but may increase the chance of a Type
II error.

3. Effect Size. Larger effect sizes are easier to detect, reducing the
probability of a Type II error.

4. Test Power (1 - β). A higher-powered test reduces the likelihood of Type

II errors and increases the test's ability to detect true effects.
 Consequences of Hypothesis Errors
[Link] I Error Consequences:
❑Implementing an ineffective policy based on incorrect research findings research
findings can drain public funds and time.
❑Approving an auditing technique or accounting method that falsely claims to
improve accuracy may lead to increased operational costs, wasted employee
hours, and incorrect financial statements.
❑Researchers advocating flawed policy decisions risk losing trust from
policymakers, stakeholders, and the public.
❑Institutions responsible for implementing ineffective programs face public
backlash, reduced political support, and long-term reputational harm.
❑An accountancy firm endorsing a flawed financial strategy risks its professional
reputation and may face lawsuits from clients.
❑Regulatory bodies relying on incorrect financial data might lose trust in the
auditing process, affecting compliance and financial market stability.
 Consequences of Hypothesis Errors
2. Type II Error Consequences:
❑Dismissing an innovative citizen feedback system that could enhance transparency results
in stagnation and missed public service improvements.
❑Ignoring a successful grassroots program because of incorrect conclusions leads to
continued inefficiencies in governance.
❑Overlooking a more effective fraud detection method due to inconclusive findings results
in ongoing undetected financial discrepancies.
❑Failing to adopt advanced analytics tools delays the organization's ability to respond to
market demands and improve forecasting accuracy.
❑Continuing with traditional practices instead of adopting successful reforms can
exacerbate societal issues like corruption, inequality, or slow economic growth.
❑Relying on ineffective disaster response policies might result in inadequate preparedness,
risking lives and property.
❑Relying on outdated or inaccurate auditing techniques might lead to undetected fraud or
errors, jeopardizing financial stability and public trust.
❑Failing to identify more efficient budgeting tools could result in poor financial planning
and wasted resources, affecting organizational sustainability.
 Strategies to Minimize Errors
[Link] Sample Size
❑Improves reliability and reduces variability in results.

2. Choose Appropriate Significance Level (α)

❑Select a level that balances the risks of Type I and Type II errors based on the context.

3. Conduct Power Analysis

❑Ensures the study has enough power to detect meaningful effects, reducing the risk
of Type II errors.

4. Use Experimental Design

❑Proper randomization, control groups, and replication improve the validity of results.
Thank You!!!