Marie Claire P.

Mandar
Plan 299 B: Research Methods

Hypothesis Testing

I. Introduction

Hypothesis testing is a decision-making process for evaluating claims about a

population. Hypothesis is defined as an assumption, an idea that is proposed for the
sake of argument so that it can be tested to see if it might be true (Merriam-Webster,
n.d.). It is like a guess or an idea that someone wants to test. Testing, meanwhile, is
defined as a critical examination, observation, or evaluation, or specifically the procedure
of submitting a statement to such conditions or operations as will lead to its proof or
disproof or to its acceptance or rejection (Merriam-Webster, n.d.). For example, if
someone thinks that "bike lanes affect traffic congestion," that is an example of a
hypothesis. Testing means putting that idea to the test by looking at the results after
trying it out.

Hypothesis testing helps us understand whether what we think might be true

actually has evidence behind it. It is widely used in research to support decision-making
based on facts, not just opinions.

II. Process Overview

Hypothesis testing is a statistical process where we examine two competing

ideas about something in the population. It usually involves two main ideas we’re testing:
the null hypothesis (H0) and the alternative hypothesis (Hα).

The goal is to see if there’s enough evidence to reject the null hypothesis in favor
of the alternative. The step-by-step process of conducting hypothesis testing is detailed
in the following subsections.

1. Start with a Question or Problem

The first step in hypothesis testing is identifying a question or problem
that you want to investigate. This question is usually inspired by something
observed or a belief that you want to test. The goal is to explore whether there’s
evidence to support or refute this idea.

2. Make a Hypothesis Example

After establishing a question, the next step is to create a hypothesis. It
typically includes:

● Null Hypothesis (H0): This states that there’s no effect or no difference.

It’s the default assumption, and we try to collect evidence to challenge it.
 ● Alternative Hypothesis (Hα): This is the statement we aim to support,
predicting that there is an effect or a difference.

A statistical hypothesis is an inference about a population parameter. This

inference may or may not be true. The only sure way of finding the truth or falsity
of a hypothesis is by examining the entire population. Because this is not always
feasible, a sample is instead examined for the purpose of drawing conclusions.

3. Collect Data
To test the hypothesis, we need to collect data that can be measured and
analyzed. This usually involves setting up an experiment or observation to gather
evidence on the effect of interest.

4. Analyze Results
The analysis step involves calculating results from the collected data and
interpreting what they mean. We often use statistical tests to help us determine if
the results show a significant effect.

5. Draw a Conclusion
Based on the analysis, we conclude whether the hypothesis was
supported by the data. We either reject the null hypothesis (indicating evidence
for an effect) or fail to reject it (indicating insufficient evidence for an effect).

III. Common Types of Hypothesis Tests

Various statistical tests are used depending on the type of data, sample size, and
research objectives. Some of the commonly used methods are the z-test, t-test, chi-
square test, and the ANOVA or the analysis of variance. The z-test is used for large
samples (n > 30), especially when we know the population’s variance. It checks if the
sample mean differs significantly from the population mean. Meanwhile, the t-test is
often used for smaller samples (n < 30) when the population variance is unknown. It’s
used to compare the means of two groups.

The chi-square test is used for categorical data to see if there’s a relationship
between two variables. It tests if observed frequencies differ from expected frequencies
in one or more categories. Finally, ANOVA is used to compare the means of three or
more groups to see if at least one differs significantly from the others. ANOVA
determines if at least one group mean is different from the others, though it doesn’t
specify which group differs.

Each of these hypothesis tests is chosen based on the type of data, sample size,
and specific question being investigated. Together, they provide researchers with a
variety of tools to test different types of hypotheses accurately and reliably.
IV. Understanding Errors in Hypothesis Testing

In hypothesis testing, there are four possible outcomes, as shown in Table 1. In

reality, the null hypothesis may or may not be true. We use data to decide if we can
reject the null hypothesis. If data shows strong evidence, we reject the null hypothesis
(meaning our alternative hypothesis might be correct). If data shows weak evidence, we
keep the null hypothesis (meaning our hypothesis might be incorrect).

Table 1. Possible outcomes in hypothesis testing

Type I Error (False Positive or False Alarm) occurs when the null hypothesis is
rejected despite being true. It represents a “false alarm” and is controlled by the
significance level (α). On the other hand, Type II Error (False Negative or Missed Alarm)
happens when we fail to reject the null hypothesis even though it is false. It indicates a
missed opportunity to identify an actual effect, and the probability of this error is denoted
by β.

The decision to reject or accept the null hypothesis is made based on the basis
of probabilities. If there is a large difference between the value of the parameter obtained
from the sample and the hypothesized parameter, the null hypothesis is probably not
true. But how large a difference is necessary to reject the null hypothesis?

V. Importance of Significance Level, P-Value, and Critical Value

Three fundamental components guide the decision-making process in hypothesis

testing, the significance level (α), P-value, and the critical value. Significance level is the
threshold for rejecting the null hypothesis, typically set at 0.05 (5%). A lower significance
level reduces the chance of a Type I error but requires stronger evidence to reject the
null hypothesis.

P-value represents the probability of obtaining results as extreme as those

observed if the null hypothesis is true. If the p-value is below the significance level, we
reject the null hypothesis. Balancing the errors is crucial. For instance, reducing the
 significance level decreases the likelihood of a Type I error but may increase the risk of
a Type II error.

The critical value is a cutoff point that helps us decide if our test results are
strong enough to say there’s an effect. It’s based on the level of significance (α) that we
choose. For example, if we set our level of significance to 5% (or 0.05), we find a critical
value that leaves only a 5% chance of making a Type I Error if we reject the null
hypothesis.

There are three types of test used to determine how we assess the significance
of our findings in relation to the null hypothesis.

1. Left-tailed test - used when the alternative hypothesis states that the
parameter of interest is less than the value specified in the null
hypothesis.
2. Right-tailed test - applied when the alternative hypothesis states that the
parameter is greater than the value specified in the null hypothesis.
3. Two-tailed test - used when the alternative hypothesis states that the
parameter is not equal to the value specified in the null hypothesis.

For example, in a two-tailed test with a 5% significance level, the critical value
might be ±1.96. If the test statistic is above ± 1.96, we reject the null hypothesis. The
critical value depends on the significance level and the type of test (one-tailed or two-
tailed).

VI. Understanding the Z-Score in Hypothesis Testing

A z-score is a standardized measure that shows how far a particular data point is
from the mean of the data set, measured in terms of standard deviations. In hypothesis
testing, the z-score helps us understand whether the difference between our sample and
population is likely due to random chance or if it’s statistically significant.
 The z-score is calculated using the following formula:

z=( χ−μ)÷ ¿

where:
● χ is the value of the sample mean,
● μ is the population mean,
● σ is the population standard deviation (if known), and
● n is the sample size.

The result tells us how many standard deviations a data point is from the
population mean. A z-score of 0 indicates the sample mean is exactly the same as the
population mean. Positive z-scores indicate values above the population mean and
negative z-scores indicate values below the population mean.

In hypothesis testing, the z-score is especially important in z-tests when we want

to determine if there’s a significant difference between a sample mean and a population
mean. The calculated z-score is compared to a critical value, which is based on the
significance level (usually 0.05) chosen for the test. The critical value for a 5%
significance level in a two-tailed test is typically ±1.96. If the z-score is beyond this
critical value, we reject the null hypothesis.

VII. Hypothesis Testing Example

In urban planning, understanding the relationship between infrastructure and

traffic dynamics is crucial. This hypothesis testing example explores whether the
implementation of bike lanes significantly affects traffic congestion levels.

Step 1: Start with a Question or Problem

● Do Bike Lanes Affect Traffic Congestion?

Step 2: Make a Hypothesis Example

● Null Hypothesis (H0): Bike lanes have no effect on traffic congestion.
● Alternative Hypothesis (Hα): Bike lanes reduce traffic congestion.

Step 3: Collect Data

Before collecting data, we define what our population and sample population are.
In this case, the population consists of all barangays in the Philippines with bike lanes
while the sample includes 100 barangays in various cities across the country where bike
lanes were implemented.

Assuming that traffic congestion is measured using the average traffic volume,
we collect data for average traffic volume before and after bike lanes are installed:
 Population data:
● Average traffic volume: 500 cars per day
● Standard deviation: 50 cars

Sample data:
● Sample size: 100 barangays
● Average traffic volume: 470 cars per day

Step 4: Analyze Results

After collecting data, we select a significance level (α) to be used for this study,
which is 0.05 or 95%. To compare the means of two independent samples using a z-
test, we will use the formula for the z-test statistic:

z=( χ−μ)÷ ¿

Substituting the values, we get a z-value of -6. For a left-tailed z-test at a

significance level (α) of 0.05, the critical z-value from the z-table is approximately -1.645
(since we are testing if the traffic volume decreases).

Step 5: Draw a Conclusion

Since -6 < -1.645, we reject the null hypothesis. The results indicate that the bike
lanes have a statistically significant effect on reducing traffic volume, as the average
traffic volume decreased from 500 cars per day to 470 cars per day. Based on these
findings, urban planners should consider expanding the bike lane network as a viable
strategy to alleviate traffic volume. This data supports the investment in cycling
infrastructure, aligning with sustainable urban mobility goals.

VIII. Conclusion

Hypothesis testing is used in many fields to make informed decisions. It is used

in medicine to test whether a new drug is effective, in the social sciences to check if a
 program or intervention has an impact, in business to see if a new product feature
improves customer satisfaction, and in urban planning to evaluate the effectiveness of
policies, programs, and interventions.

It is important to note that hypothesis testing also has its limits. It doesn’t prove
whether a hypothesis is right or wrong, it does however show if there is enough evidence
to support or reject it. Furthermore, results can depend on the quality and size of the
sample. Poor data also plays a huge role in hypothesis testing as it can lead to
unreliable conclusions.

Hypothesis testing is a valuable tool in research for checking if ideas are

supported by evidence. By following a structured process, researchers can make
informed decisions based on data rather than guesses. While hypothesis testing isn’t
perfect, it gives a reliable framework to explore and understand different ideas.

IX. References:

Alferez, M. and Duro, M.C. (2023). Statistics and Probability. MSA Academic Advancement
Institute. MSA Publishing House. ISBN 971-8740-82-1.

Merriam-Webster. (n.d.). Hypothesis. In Merriam-Webster.com dictionary. Retrieved November

3, 2024, from https://www.merriam-webster.com/dictionary/hypothesis

Merriam-Webster. (n.d.). Test. In Merriam-Webster.com dictionary. Retrieved November 3,

2024, from https://www.merriam-webster.com/dictionary/test