See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/378210052

Quantitative Methods in Social Science Research: Systematic Review of Content

Analysis, Survey and Experiment Methodologies

Preprint in SSRN Electronic Journal · February 2024

DOI: 10.2139/ssrn.4698175

CITATIONS READS

3 8,436

1 author:

Kazi Mehedi Hasan

Southern Illinois University Carbondale
4 PUBLICATIONS 14 CITATIONS

SEE PROFILE

All content following this page was uploaded by Kazi Mehedi Hasan on 15 February 2024.

The user has requested enhancement of the downloaded file.

Quantitative Methods in Social Science Research: Systematic Review
of Content Analysis, Survey and Experiment Methodologies

Kazi Mehedi Hasan

PhD Candidate, Mass Communication
Southern Illinois University, Carbondale, IL

Abstract

Quantitative research methods aim to be objective and unbiased; this

means that the data collection and analysis process should be designed to
minimize the influence of the researcher's personal opinions or biases.
Therefore, data collection methods should be reliable, meaning the same
results would be obtained if the study were repeated under the same
conditions. To do this, researchers follow some general rules of data
gathering such as exact process (survey, experiment, observation &
content analysis), defined units, and variables. Quantitative research
results can be generalized so that the study can be applied to a larger
population than where it was studied. These methods are widely used in
various fields, including natural science, science, education, business, and
marketing campaigns. Each method's potential opportunities, strengths,
and limitations are discussed.

Introduction
Quantitative research systematically investigates social phenomena using
statistical or numerical data (Watson, 2015). In other words, it is a research procedure
that collects, analyzes, and interprets numeric data (Davis, 2011). For example,
quantitative research collects data by sending online surveys, online polls, and
questionnaires to existing and potential customers using sampling methods. What types
of information the media transmit and what effect they have on society are found
through quantitative research. For instance, information the media is passing on can be
gathered through content analysis, and the impact can be measured by surveying media
users.

Social science research aims to explore and comprehend human behavior and
relationships. Different aspects of human behavior require investigation to gain insights
into their perspectives and responses toward the social world. Generally, quantitative
research methodologies involve social science research gathering and analyzing
numerical data. Observation, survey, experiment, and content analysis are the most
used methodologies in quantitative research, also known as primary research methods.

Quantitative research focuses on the specific behavior that can be quantified. It

includes a range of related methods subject to statistical data analysis. Some of these
phenomena are easily measurable, such as age and height. However, experiences and
feelings are not easily measurable. Quantitative research encompasses everything in
which specific methods and criteria are applied to measure difficult issues. Quantitative
research is essentially data-oriented. Based on data availability, this research is mainly
divided into two parts. They are Primary Research and Secondary Research.
A. Secondary Research
Secondary research involves collecting and analyzing information and data that other
researchers or organizations have gathered (Johnston, 2014). This type of research
involves reviewing and synthesizing existing literature, reports, and data sources to
answer a research question or address a specific problem. Secondary research is often
used as a starting point for a research project, as it can provide a foundation of
knowledge and context for the topic being studied (Hasan, 2023; Ramdhani et al.,
2014). It is also useful when conducting research that may be difficult, time-consuming,
or expensive to conduct from scratch, such as large-scale surveys or experiments. There
are two important reasons for secondary research (Kolb, 2012).
1. Start new research: When researchers look for a problem definition, they use
secondary data to start the project. Secondary researchers have no control over
secondary data.
2. Institutional Development: The government, institution, company, media, or
any other sector needs to evaluate its own program and activities. Secondary
research sources include academic journals, government reports, industry
publications, market research reports, and online databases (Kolb, 2012).

B. Primary Research
Primary research is designed to meet specific needs and obtain desired information. It
also involves collecting information from research participants through surveys or
asking questions. Currently, four types of primary research methods are used in the
social science.
Quantitative Research methods are:
1. Observation
2. Content analysis
3. Surveys
4. Experiments

This study systematically reviews research papers related to quantitative research

methods to identify the most common definition used by researchers to find unbiased
ways of numeric data collection. It helps researchers discover all related data gathering
structures and procedures in a single place so they can easily follow organized
measures. It also discusses the strengths and weaknesses of all methods and techniques
so that researchers can look for alternative ways of minimizing bias and potential harm
getting results that represent reality and are easily generalizable.
 1. Observation
Observational research can be either quantitative or qualitative. It is a research
technique where researchers observe the ongoing behavior of participants in natural or
normal situations. It is also an important and complex research method (Ciesielska et
al., 2018). The purpose of such research is to gather more reliable insights.
Observational research is done in two ways: Human observation and automated
observation (Davis, 2011).
Human Observation: The researcher observes other participants, whereas the
researcher inserts himself into the environment. Sometimes, the researcher does not
intervene in the setting. In that case, researchers monitor the participants remotely or in
a laboratory setting.
Human observation is most appropriate in four situations (Davis, 2011). It includes
where observation of behavior is more insightful than description when attitudes are
difficult to verbalize, when survey methods do not accurately represent human
behavior, and when behaviors themselves are the best source of insight. Human
observation research is especially done in scientific observation, physiological findings,
and market or product research. The researcher must be able to interpret the
observation into quantity (quantitative) or pattern/category (qualitative). Human
observation research relies on four aspects (Davis, 2011).
 Situation: Natural vs. Artificial
 Observer obtrusiveness: Open vs. Disguise
 Observer participation: Active vs. Passive
 Data recording: Structured vs. Unstructured
Natural settings: Largely in the social sciences, natural observational investigation
refers to the study conducted outside the laboratory to observe the behavior, particularly
in participants' own environment where they act naturally (Crano et al., 2011). For
example, observing customers in a restaurant examine their behavior according to their
gender and race, observing sales clerks' behavior in different outlets as they serve
customers, and observing media watching and newspaper or magazine reading in the
doctor's waiting room. It must have three components; natural situation, natural event,
and natural behavior.
There are some important strengths of natural settings observational research; it
includes interpretation of complex and novel situations, useful measuring authentic
behavior, and Development of new theories. Limitations include the researcher having
no control over the situation and not being useful for all types of observation. It is also
difficult to record or observe all happenings in a natural setting (Crano et al., 2011).
Artificial settings: An artificial observational setting refers to the behavior of
participants being tested in a laboratory or controlled environment. For example, the
traditional media and social media using participants' behavior can be examined in a
specific room or laboratory setting. Artificial observation involves creating fabricated
situations and manipulating content to record behavior (Davis, 2011).
 However, artificial observation has two advantages over natural observation. It
speeds up data recording and gathering. It can also control external factors in the
research.
Open and Disguised Observation: In the case of open observation, the
researcher is known, and participants are informed that their activities in an event are
observed, followed, and data is collected for the research purpose (Davis, 2011).
Disguised observation is a special type of market research technique where
research participants are not aware that they are being observed and the researcher's
presence is unknown (Gardner, 2000; Boote & Mathews, 1999). Disguised observation
generally provides more realistic data about customer/ participant behavior than open
observation.
Active and Passive Observation: An active observer, as does an open observer,
observes the research participant directly. Like a disguised observer, a passive observer
collects data on the participant's behavior without their knowledge (Cohen et al., 2017).
Both observation processes have advantages and disadvantages; an active
observer acts as a part of observation and direct data collection, and they have the
opportunity to ask questions and interact with the participant. However, passive
observers have no chance to interact or ask questions. An active observer has the
potential to introduce bias, but a passive observer has not. This process is time-
consuming but collects more clean and authentic behavioral data than active
observation.
Structured and Unstructured
The structured form of the data collection process relies on a prior checklist
before observation. Researchers need to fix where they would focus on. Mostly,
quantitative research records data in a structured way. Structured observation requires
more time to prepare checklists before research, more time to collect data, and less time
to analyze because everything is structured.
Unstructured data collection provides greater opportunities for new knowledge
through observation. It records in a written narrative form so the observer can do so
easily, but data coding and data analysis are time-consuming.
Automated observation
Automated observation, where participants' behavior is monitored using
technology or mechanical devices rather than human observers (Davis, 2011). It has
three forms: Direct monitoring (online/offline), Observing Consumer-Generated
Content, and eye tracking.
A website can provide the information it needs and adapt to the needs of its
users by monitoring, storing, and analyzing the interaction data of its users (Zdziebko &
Sulikowski, 2015). For example, Facebook, Google, or YouTube analyze the users'
data and optimize the posts and videos according to their preferences. These are done
directly in the observing consumer-generated content process. Websites automatically
deliver product advertisements to consumers through data observation.
Eye Tracking
Eye tracking is an experimental method where an internal device of a product or
advertisement automatically records gaze position and eye motion across the time of
observation (Carter & Luke, 2020). Nowadays, eye-tracking technology has improved,
which can record the natural activities of the respondents (Davis, 2011). Eye tracking is
commonly used in online and offline advertising to track customer interest and
rearrange it to their preferences.

2. Content analysis
Content analysis is a systematic method of content evaluation concerning the
meanings, contexts, and intentions contained in messages (Morris, 1994). It is a
systemic evaluation process of identifying the communication text's embedded
meaning, pattern, and phenomenon (Hasan, 2020). Content analysis is an analytical
tool for analyzing and monitoring documents or content for the researchers and the
"most important research technique in the social science" (Krippendorff, 1989, p.403).
Definition
Content analysis is generally described as a scientific method for analyzing
media and communication messages, forming the basis for drawing hypotheses and
conclusions about the content. Krippendorff (1989) also defines "content analysis as a
research technique for making replicable and valid inferences from data to their concept
(p-403)". It systematically narrows down and compresses documents into specific types,
categories, and numeric data.
Scopes, Areas, Procedure
Content analysis research has large scopes and potentiality in social science
research for its accommodation and incorporation characteristics. The content analysis
technique is not limited to the domain of communication text analysis but can be
applied to other areas (Stemler, 2000). For example, newspaper or radio-TV
contentment can be applied to studies on film, audio-video, image, drawing, social
media, web, blog, and other cultural and archeological documents.
Although most commonly used in mass communication research, content
analysis methods are also used by scholars in other social science disciplines such as
communication, psychology, political science, history, and language studies (Prasad,
2008). In particular, it has been used extensively to explore the nature of news coverage
of social issues and to understand broad themes such as social change, cultural symbols,
social change, and changes in media content over time. This analysis discusses the
media's role in different national and international issues, media treatment towards
women or minorities, and electoral issues reflected in media content (Prasad, 2008).
Content analysis, like any other research method, follows certain general conditions.
Those three are objectivity, systematic, and generalization (Prasad, 2008).
1. Objectivity: Certain rules must be followed in the case of qualitative analysis
so that multiple researchers can obtain the same results from the same content. For this
reason, the reliability of the data is ensured through multiple coders while data coding.
No, bias, data should be valid sources, and procedures should be reliable.
2. Systematic: This system requires inclusion or exclusion from the rules.
However, the material and bias supporting the researcher's hypothesis must be
excluded. For this reason, although the researcher has freedom while selecting the
sample, it must be logical and methodically flawless.
3. Generalization: Quantitative research is conducted based on specific
theories. As a result, the research results can be considered effective and useful for
similar situations. Krippendorff (1989) illustrates six procedures and criteria researchers
must consider while conducting content analysis. These are designing, sampling,
unitizing, coding, data validation, and drawing inferences. Inductive and deductive
data coding processes can use content analysis, it depends on the purpose of the study.
In inductive analysis, researchers follow open coding to group and categories the
content and then analyze the data. In contrast, deductive analysis first develops
structured data analysis matrices, then defines categories and performs data coding
according to categories. Hypothesis testing and results are based on categorized data.
Debi Prashad (2008) discusses a few important strengths and weaknesses of content
analysis in his article. The strengths and weaknesses of content analysis are:
Strengths
a. It can make a quantitative expression of phenomena, expressing them in numbers
and percentages, which are more specific and objective. Analyze and interpret existing
embedded data in media and other texts.
b. It is an unobtrusive research technique that allows for studying sensitive research
topics.
c. It is context-sensitive and, therefore can process the symbolic meaning of the data.
Although it is seen as a predominantly quantitative method, it can also effectively
capture qualitative content.
d. It provides a safe method and objective data because if the researcher finds wrong
coding somewhere in the study or something is left out, that mistake can be easily
corrected.
e. It can handle a large volume of data. Although processing or coding is difficult, using
computers and software has made data processing and analysis easier.
Weaknesses
1. The assumptions of this method are limited to the content of the text only. The
researcher follows coding based on constructed theories or their own definitions. As a
result, the researcher may not decode the message's communicative meaning.
2. The results may not do justice to the content regarding semantic differences in words.
3. It often stops analysis by counting frequencies, which may fail to capture the
significance of these words.
4. Reliability and validity issues in content analysis remain unresolved despite caution.
5. The method cannot be used to test causal relationships between variables.

3. Survey Method
The survey research method is a data collection procedure. The survey procedure
provides an at-a-glance view of respondents about their attitudes, beliefs, and behaviors
(Davis, 2011). Surveys are usually conducted using questionnaires that ask questions of
interest, providing insight into specific areas. A survey method is a systematic process
or technique in which the researcher collects data for the research by asking questions
to a predetermined population or a sample group (Hasan, 2023). According to
Pinsonneault and Kraemer (1993), "means for gathering information about the
characteristics, actions, or opinions of a large group of people" (p. 77). Owens (2000)
posits some unique characteristics of the survey method; it gathers information from
primary sources that are never possible from other sources. Mostly ensures an unbiased
and representative sample population. Each respondent is asked to answer the same
question. Survey data is also used for data verification of secondary data (Owens,
2000). Kraemer (1991) specified three distinct characteristics of the survey research
method. First, survey research aims to quantify specific characteristics of a given
population. These features usually determine and test the relationship between
variables. Second, the data required for survey research is collected from people and,
therefore, can be subjective. Finally, survey research is conducted on a selected
population segment (sample), and then the results can be generalized to the whole
population for a similar context.
Generally, the researcher has no explicit control in survey research, defined
independent and dependent variables run it. However, the researcher has some sort of
implicit control. They can predict the expected relationship among the variables. The
researcher first constructs a model and then tests it by surveying (Glasow, 2005).
There are four data collection modes in the survey method:
 Personal or Face to Face (Intercept/ Pre-Recruited)
 Telephone Interviews
 Mail Surveys
 Website/online
Personal or Face to Face (Intercept/ Pre-Recruited)
Two types of face-to-face surveys are conducted. First, the intercept, where the
potential interviewee is stopped abruptly and induced to participate in the survey.
Shopping malls, airports, and stations are good places to conduct intercept surveys. The
advantage is that there is no confirmation bias in the researcher. The disadvantages are
that people often refuse to participate in surveys, and many random participants have
no experience with the research topic.
Another Face to Face survey is the pre-recruited or pre-planned interview. For
this survey of interviews, respondents are contacted in advance, interested individuals
are asked to participate, and the researcher/ data collector and participants mutually
agree on a convenient time for the interview or survey. Its advantages are that
researchers get high-quality data with a low refusal rate. The researcher can select
percipients, whatever they need for collecting information. For example, in order to
research social media platforms' security, researchers may conduct Facebook or
Twitter's responsible officers. The disadvantage is that the researcher may be biased in
selecting participants.
Face-to-face or Personal interview ensures high-quality data because the
researcher can take advantage of the presence resulting in a high response. They have
the opportunity to ask comprehensive questions. However, the disadvantage is that the
personal interview process is time-consuming, often expensive, and hard to obtain true
random samples (Davis, 2011).
Telephone Interviews
Interviewing by telephone makes the researcher's job much easier. The
researchers can collect data from their own locations. It ensures quality data at a low
cost and in less time, and random samples are easily obtained through telephone
interviews but need training for the interviewers. Network breakdown can be a huge
problem for telephone interviews, and sometimes communication gap happens between
the interviewer and participants.
Computerized telephone interviews have accelerated data collection, but
researchers have to concern about the high cost. However, researchers have found that
technology-based telephone interviews provide more data in less time, but the high
refusal rate is also a big concern (Davis, 2011; Kempf & Remington, 2007).
Mail Surveys
Until the beginning of this century, mail surveys were the data collection tool of
various companies and researchers. This method sends stamped return envelopes and
questionnaires for the survey. Sometimes participants are offered incentives (gift cards,
coupons) for participating in the survey. Respondents can easily fill out questionnaires
at their convenience, and mailing cost is adjustable to other survey procedures (Davis.
2011). However, the application of mail surveys has decreased to a large extent due to
the Internet facility. It ensures quality data at a low cost without physically interacting
with the data collector. A disadvantage is that people rarely complete the survey
questionnaire, which is time-consuming. It is not feasible in the present modern reality.
Web (Online/offline) interview
The widespread use of the Internet in the current era has increased both the
possibilities and scope of web or online surveys. Especially with internet connection on
mobile phones, it has become easier to get data through online surveys. Surveys can
easily be taken online, allowing international participants to participate and generating
large sample sizes. A major limitation of this survey is the exclusion of marginalized
groups to whom the Internet does not reach.

4. Experiments
The experimental method is a frequently used research design in social science to
examine causal relationships between variables. In this method, researchers manipulate
an independent variable and observe its impact on a dependent variable while
controlling for other variables that may influence the outcome. The goal is to determine
the cause-and-effect relationship between the studied variables and how changes in one
variable affect one or more other variables (Davis, 2011).
Surveys vs. Experiment: Though both survey and experimental research involve
data collection from the respondents, they differ (Davis, 2011). Survey research involves
collecting data through questionnaires, interviews, or other self-reporting forms. Survey
research is time-oriented, aiming to gather information about people's attitudes, beliefs,
behaviors, or experiences at a particular time. Survey research can be conducted
through various methods, such as online surveys, phone surveys, or face-to-face
interviews. In the survey, respondents can express which ad banner is relevant to their
interests.
On the other hand, experimental research goes beyond ad characteristics.
Manipulation of the research allows respondents to determine how much the banner ad
is relevant (high or low) to them (Davis, 2011). The goal is to establish cause-and-effect
relationships between variables. Experiments are often conducted in a controlled
setting, and participants are randomly assigned to experimental and control groups.
Experiment research involves the manipulation of an independent variable to observe
the effect on a dependent variable while controlling for other variables that may affect
the outcome.
Joel Davis (2011) mentions four basic steps of every experiment research regardless
of subject matter and research context. They are:
 Identify what you need to learn (Research purpose)
 Take relevant actions (Manipulations)
 Observe effects and consequences (Collecting data)
 Determine causality (Analyzing data) (Davis, 2011, p.364)

Independent and Dependent Variables

To conduct experiment research, the researcher needs at least one independent
variable, one dependent variable, and manipulation of one independent variable.
Experimental research examines whether the independent variable is responsible for
changes in the dependent variable. For example, to test whether political advertising
persuades voters to vote, one group of voters is shown one candidate's ad while the
other is not. Then, the two groups of voters are asked whether they will vote for that
candidate. Joel Davis (2011) argues that the researcher needs to follow some
requirements to claim causality between the variables.
a. Events must occur in proper order: To the causal relationship using stimuli or
manipulation should be in the same order. If a researcher wants to show more than one
ad, she/he must follow the same order for all participants.
b. Events must occur simultaneously and show an explicit relationship: To maintain
concomitant variation, the researcher needs to use certain factors at the same time.
c. Alternative explanations must be reduced or eliminated: Participants must be
shown the same manipulation because different manipulations can produce different
results. All voters in a group should be presented with advertisements of equal length;
no difference can be shown in the same group. Furthermore, a consistent medium
should be utilized for the manipulation; the same advertisements cannot be shown in a
different medium in the same group. For example, shown by someone on TV or a
computer or mobile phone in a group because it can create an alternative causal result.
d. relationship must show the strength of association: The researcher must solve
internal validity problems. Internal validity threats include problems associated with an
initial pre-test survey, problems due to data collection, Problems associated with the
sample selection, Problems caused by the study context and history, and Problems
attributable to researcher bias (Davis, 2011, p.367).
Researchers conduct various types of experimental design research.
Experimental designs vary in structure, using the number of variables and methods for
data collection. Here are two common types of experimental designs:
 Quasi-experimental Designs
 True Experimental Designs
Quasi-experimental Designs
The quasi-experimental designs attempt to uncover a causal relationship. Still, they
are not considered true experiments because it has no control group and does not use a
random selection of the sample. Their inability to eliminate numerous threats to
internal validity in determining causality (Davis, 2011). The three most common types
of quasi-experimental designs are:
 One group post-test only,
 One group pre-tests to pro-test,
 Two group post-test with control. (Davis, 2011, p.372)
One group post-test only is called a weak experiment design because it has no control
group and follows a convenient sample. It takes one group to post-test and exposes
them to the experimental manipulation. This design is unable to control internal
validity threats, and data cannot be generalizable.
One group pre-tests to pro-test: In this design, a single group is selected through
convenience sampling. Data is initially gathered from the participants, then
manipulated by the independent variable, and then repeated data collection. The main
disadvantage of this design is no control group that threatens internal data validity.
Two-group post-test with control: The design involves testing between two groups.
The first group is subjected to pre-test and post-test data collection, while the second
group only undergoes post-test data collection. Any changes in behavior between the
two groups are then observed and compared. The main problem of this design is the
group characteristics; there is no way to make sure the two groups are equivalent, and it
is difficult for researchers to determine whether the outcome is a result of the treatment
or the characteristics of the sample.
True Experimental Designs
There are two significant differences between true experimental designs and
quasi-experimental designs. First, true experimental designs have a control group.
Second, they employ random assignment to create test and control groups, ensuring
that both groups are assumed to be equivalent at the start of the test. This approach
guarantees that each participant has an equal chance of being assigned to either the
control or treatment group (Davis, 2011).
There are five types of true experimental designs.
I. Simulated pre-test to post-test (Two groups – Two measure design)
II. Post-test only with control (Two groups – Two measure design)
III. Pre-test to post-test with control (Two groups with Four measures)
IV. Solomon four-group design (Four groups – Six measures)
V. Factorial designs (More than one manipulation
Two groups – Two Measure Design
A two-group, two-measure design is the most common form of experimental
design that involves comparing two groups on two different measures. In this design,
two groups of participants are randomly assigned to either a treatment or control
condition, and both groups are measured on two different variables at two different
times. This design contains two types of true experimental design, simulated pre-test to
post-test and post-test only with control.
Simulated pre-test to post-test: A simulated pre-test to post-test design is a
research design involving two groups of participants where half are randomly assigned
a pre-test to evaluate before manipulation, and the rest half are randomly assigned a
post-test to evaluate after manipulation. The advantage of this design is that it
effectively addresses concerns related to maturation, instrumentation, pre-
measurement, and interaction threats to internal validity. However, it has a limitation
in that it does not account for the effect of historical events that may occur during the
time between the pre-test and post-test.
Post-test only with control (different Sample/manipulation/Same time): This
experimental design also involves two randomly assigned groups for post-test
evaluation. However, the difference between the previous pre-test - and post-test design
is the time. Two different groups are assigned to evaluate two different measurements
simultaneously. For instance, one group is assigned to evaluate a pizza with low water,
and another is assigned to evaluate a pizza with sufficient water simultaneously. This
design can treat internal validity, including maturation, instrumentation, pre-
measurement, interaction, and historical factors. However, a limitation of this design is
that it does not confirm whether random assignment resulted in equivalent pre-
treatment characteristics in the treatment and control groups.
Pre-test to post-test with control (Two groups with Four measures): In this
design, researchers measure two groups of participants on four different variables both
before and after an intervention. One of the groups serves as the control, and another as
the treatment. This experimental design is commonly used to evaluate the effectiveness
of an intervention or treatment by comparing pre and post-test scores of the treatment
and control groups across multiple measures. This design eliminates selection threats.
Solomon four-group design (Four groups – Six measures): The Solomon four-
group design is the most useful and powerful experimental research design. It controls
all major threats to internal validity (Davis, 2011). This design involves four groups of
participants with six measurements. The design involves randomly assigning
participants to one of the four groups. In this process, two groups receive a pre-test, and
two groups do not receive a pre-test. This design is commonly used in experimental
research to evaluate a pre-test's effect on a study's results.
Factorial Design: Researchers can manipulate two or more independent
variables simultaneously. Each variable has two or more levels or aspects
simultaneously to examine their main effects and interactions on a dependent variable.
In a factorial design, each level of one independent variable is combined with each level
of the other, resulting in a complete matrix of all possible combinations. For example,
in a 2 x 2 factorial design, there are two independent variables, each with two levels.
There are four possible combinations or conditions in a factorial design: For example,
independent variables are price and brand, where factorial design consists of 1)
International brand and high price, 2) International brand and low price, 3) Local brand
and high price, and 4) Local brand and low price. Respondents are randomly assigned
to one of the four cells to evaluate the condition in this design.
Factorial designs offer an advantage by examining multiple independent
variables and their interactions in a single study, resulting in a more comprehensive
understanding of the research question. Therefore, factorial designs are still the most
important and frequently employed approach in experimental research.

Strengths and Limitations of Quantitative Methods

Johnson and Onwuegbuzie (2004) explain some strengths and weaknesses of
quantitative research.
Strengths
1. Tests and validates previously developed theories about how similar events
occur
 2. Tests hypotheses that are constructed before collecting data
3. Research based on random sample data can generalize the results
4. Research based on the general population and subpopulation can generalize the
results
5. Obtained data allows quantitative predictions
6. Eliminates the confounding influence of many variables and allows more
credibly established cause-and-effect relationships
7. Data collection using quantitative methods is relatively quick (e.g., telephone
interviews, close-end survey)
8. Provides precise, quantitative, numerical data
9. Data analysis is relatively less time-consuming (using statistical software)
10. The research results are relatively independent of the researcher (e.g., statistical
significance)
11. Quantitative data may have higher credibility with many people in power (e.g.,
Administrators, politicians, and people who fund programs)
Weakness
Johnson and Onwuegbuzie (2004) identify some weaknesses of quantitative research.
1. Sometimes, the categories used by the researcher are not compatible with other
locations or situations.
2. The theory used in research may not apply to particular areas or not reflect local
constituencies' understandings.
3. Researchers can miss phenomena due to confirmation bias or focus on the
wrong theories and hypotheses.
4. The conclusion and result produced by quantitative research might be too
abstract and general for direct application to specific local situations, contexts,
and individuals.

Conclusion
The main critique of quantitative research is that it may not capture the full range of
human experiences or emotions and may overlook important contextual factors that
influence behavior. However, the ability to produce objective, systematic data
collection and analysis to provide precise measurements of variables and statistically
significant results, to generalize findings to larger populations, and to facilitate
comparisons across different groups or periods allow this method advantages over the
other methods (Davis, 2011). Large-scale quantitative research may raise ethical
concerns regarding data privacy, informed consent, and potential harm to participants
(Eynon et al., 2008). Statistical analysis relies on certain assumptions about the data
and the research design, which, if violated, can lead to misleading conclusions (Myers
et al., 2013). Researchers stick to following general rules of quantitative data collection
to avoid any bias and assumptions.
 References
Boote, J., & Mathews, A. (1999). "Saying is one thing; doing is another": The role of
observation in marketing research. Qualitative Market Research: An International
Journal, 2(1), 15-21.
Carter, B. T., & Luke, S. G. (2020). Best practices in eye tracking research. International
Journal of Psychophysiology, 155, 49-62.
Ciesielska, M., Boström, K. W., & Öhlander, M. (2018). Observation methods.
In Qualitative methodologies in organization studies (pp. 33-52). Palgrave Macmillan,
Cham.
Cohen, L., Manion, L., & Morrison, K. (2017). Observation. In Research methods in
education (pp. 542-562). Routledge.
Crano, W. D., Brewer, M. B., & Lac, A. (2014). Principles and methods of social research.
Routledge.
Davis, J. J. (2011). Advertising research: theory & practice. Pearson Higher Ed.
Eynon, R., Fry, J., & Schroeder, R. (2008). The ethics of internet research. Sage internet
research methods, 23-41.
Gardner, F. (2000). Methodological issues in the direct observation of parent–child
interaction: Do observational findings reflect the natural behavior of
participants?. Clinical child and family psychology review, 3(3), 185-198.
Glasow, P. A. (2005). Fundamentals of survey research methodology. Retrieved
January, 18, 2013.
Hasan, K. M. (2020). News Framing In Bangladesh, India And British Media: Bangladesh
Parliamentary Election 2018 (Doctoral dissertation, The University of Mississippi).
Hasan, K. M. (2023). Partisanship and Media Choice on Political Information and
Issue Diversity in the Unites States. Available at SSRN 4547684.
Johnson, R. B., & Onwuegbuzie, A. J. (2004). Mixed methods research: A research
paradigm whose time has come. Educational researcher, 33(7), 14-26.
Johnston, M. P. (2014). Secondary data analysis: A method of which the time has
come. Qualitative and quantitative methods in libraries, 3(3), 619-626.
Johnston, M. P. (2014). Secondary data analysis: A method of which the time has
come. Qualitative and quantitative methods in libraries, 3(3), 619-626.
Kempf, A. M., & Remington, P. L. (2007). New challenges for telephone survey
research in the twenty-first century. Annu. Rev. Public Health, 28, 113-126.
Kolb, B. (2012). Conducting secondary research. Sage secondary data analysis, 111-128.
 Kraemer, K. L. (Ed.). (1991). The information systems research challenge (vol. III) survey
research methods. Harvard University Graduate School of Business
Administration.
Krippendorff, K. (1989). Content analysis.
Morris, R. (1994). Computerized content analysis in management research: A
demonstration of advantages & limitations. Journal of management, 20(4), 903-
931.
Myers, J. L., Well, A. D., & Lorch Jr, R. F. (2013). Research design and statistical analysis.
Routledge.
Owens, L. K. (2002). Introduction to survey research design. In SRL fall 2002 seminar
series (Vol. 1).
Pinsonneault, A., & Kraemer, K. (1993). Survey research methodology in management
information systems: an assessment. Journal of management information
systems, 10(2), 75-105.
Prasad, B. D. (2008). Content analysis. Research methods for social work, 5, 1-20
Ramdhani, A., Ramdhani, M. A., & Amin, A. S. (2014). Writing a Literature Review
Research Paper: A step-by-step approach. International Journal of Basic and Applied
Science, 3(1), 47-56.
Stemler, S. (2000). An overview of content analysis. Practical assessment, research, and
evaluation, 7(1), 17.
Watson, R. (2015). Quantitative research. Nursing standard, 29(31).
Zdziebko, T., & Sulikowski, P. (2015). Monitoring human website interactions for
online stores. In New Contributions in Information Systems and Technologies: Volume
2 (pp. 375-384). Springer International Publishing.

View publication stats