This document provides an overview of experimental design in research. It defines experimental design as a scientific approach to research where independent variables are manipulated to measure their effect on dependent variables. The key types of experimental design are described as pre-experimental, quasi-experimental, and true experimental. True experiments offer the strongest causal inferences because variables are manipulated and subjects are randomly assigned, allowing for comparison. Experimental research is described as useful for businesses to test new strategies, understand customer needs, and assess the business environment before making changes. The steps for conducting experimental research are outlined.
1. MODULE CODE & TITLE: MSCM704 RESEARCH METHODS
NAME SURANAME REG NUMBER
1.BLESSING C MBIRIVIRI R2216752G
MODE OF ENTRY: HARARE WEEKEND CLASS
LEVEL: 1:1
PROGRAMME: MASTER OF COMMERCE IN SUPPLY CHAIN
MANAGEMENT
LECTURER: DR W Dzimiri
2. EXPERIMENTAL DESIGN
INTRODUCTION
Conducting credible and trustworthy research to inform managerial
decisions is arguably the primary goal of business and management
research. Research design, particularly the various types of experimental
designs available, are important building blocks for advancing toward this
goal. Key criteria for evaluating research studies are internal validity (the
ability to demonstrate causality), statistical conclusion validity (drawing
correct conclusions from data), construct validity (the extent to which a
study captures the phenomenon of interest), and external validity (the
generalizability of results to other contexts). Perhaps most important,
internal validity depends on the research design’s ability to establish that
the hypothesized cause and outcome are correlated, that variation in them
occurs in the correct temporal order, and that alternative explanations of
that relationship can be ruled out.
Research designs vary greatly, especially in their internal validity.
Generally, experiments offer the strongest causal inference, because the
causal variables of interest are manipulated by the researchers, and
because random assignment makes subjects comparable, such that the
sources of variation in the variables of interest can be well identified.
Natural experiments can exhibit similar internal validity to the extent that
researchers are able to exploit exogenous events creating (quasi-
)randomized interventions. When randomization is not available, quasi-
experiments aim at approximating experiments by making subjects as
comparable as possible based on the best available information. Finally,
non-experiments, which are often the only option in business and
management research, can still offer useful insights, particularly when
changes in the variables of interest can be modelled by adopting
longitudinal designs.
DEFINITION OF KEY TERMS
1. EXPERIMENTAL DESIGN - is the process of researching in an
objective and controlled manner to optimize precision and reach
conclusions about a hypothesis statement. The goal is to determine
3. the effect a factor or independent variable has on a dependent
variable.
2. EXPERIMENTAL DESIGN- is a scientific approach to research,
where one or more independent variables are manipulated and
applied to one or more dependent variables to measure their effect
on the latter. The effect of the independent variables on the
dependent variables is usually observed and recorded over some
time, to aid researchers in drawing a reasonable conclusion
regarding the relationship between these 2 variable types.
EXPERIMENTAL DESIGN OVERVIEW
Experimental research is a type of scientific examination in which one or
more independent variables are changed and then applied to one or more
dependent variables to see how they affect the latter. The effect of
independent variables on dependent variables is frequently observed and
recorded over time to help researchers reach a plausible conclusion about
the link between these two types of variables. The experimental research
approach is frequently employed in the physical and social sciences,
psychology, and education. It is based on a simple logic that compares
two or more groups, but it can be challenging to implement. Experimental
research designs, most associated with laboratory test procedures, entail
gathering quantitative data and doing statistical analysis on it during the
study process.
The Key Aspects of Experimental Research
There are various attributes that are formative of and unique to
experimental research in addition to its main purpose. Understanding
these is key to understanding this kind of research in-depth and what to
expect when performing it.
The following enumerates the defining characteristics of this kind of
research:
It includes a hypothesis, a variable that will be manipulated by the
researcher along with the variable that will be measured and
compared.
The data in this research must be able to be quantified.
The observation of the subjects, however, must be executed
qualitatively.
4. It can be conducted in a laboratory in field settings, i.e., field
research.
The latter is rarer, as it is difficult to manipulate treatments and to
control external occurrences in a live setting.
It relies on making comparisons between two or more groups (the
variables).
Some variables are given an experimental stimulus called a
treatment; this is the treatment group.
The variables that do not receive a stimulus are known as the control
group.
First, researchers must consider how the variables are related and
only afterward can they move on to making predictions that can be
tested.
Time is a crucial component when putting forth a cause-and-effect
relationship.
There 3 types of experimental research:
Pre-experimental research design
True experimental research design
Quasi-experimental research design
Types of experimental design
The types of experimental study designs are into three types as pre-
experimental, quasi-experimental, and real experimental.
5. Pre-experimental research design:
This entails a group or several groups to be observed after factors of
cause and effect are implemented. Researchers implement this research
design when they need to learn whether further investigation is required
for these groups.
Pre-experimental research has its own three subtypes:
One-shot Case Study Research Design
One-group Pre-test-post-test Research Design
Static-group Comparison
Quasi-experimental Research Design
Representing half or pseudo, the moniker “quasi” is used to allude to
resembling true experimental research, but not entirely. The participants
are not randomly assigned, rather they are used when randomization is
impossible or impractical. Quasi-experimental research is typically used
in the education field. Examples include: the time series, no equivalent
control group design, and the counterbalanced design.
6. True Experimental Research Design
This kind of experimental research design studies statistical analysis to
confirm or debunk a hypothesis. It is regarded as the most accurate form
of research. True experimental research can produce a cause-effect
relationship within a group.
This experiment requires the fulfilment of 3 components:
A control group (unaltered) and an experimental group (to undergo
changes in variables)
Random distribution
Variables can be manipulated
Why Your Business Needs Experimental Research
There are various benefits to conducting experimental research for
businesses.
Test New Business strategy
Firstly, this form of research can help businesses test a new strategy
before fully engaging in/ launching it. The strategy can involve anything
from content marketing strategy to a new product launch. This is
especially useful for technology companies, which conduct
experimentation frequently. In fact, this kind of research is essential to an
R & D (research and development) department. This makes experimental
research a much-needed effort when it comes to spurring innovation.
Whether it involves a slight rebranding or an upgrade of products,
experimental research guides these campaigns in a science-backed
manner.
To meet customer needs
Secondly, a business must excel in meeting customer needs. Customer
experience is an overwhelmingly important side of any business, as
customers are willing to make on-the-stop purchases and pay more for a
good CX. As such, each product addition and change in a customer
journey must be carried out wisely. Businesses ought to avoid creating
unwanted services, or those that cause any aversion within customers.
Instead, they should only invest in the most profitable services, products
and experiences, a feat that cannot be accomplished solely on
guesswork. Experimenting allows brands to understand customer
7. preferences and changes in their behaviours, as the experiments create
stimuli and changes in independent variables.
Business environment Assessment
Additionally, experimental research grants companion an understanding
of their business environment. In turn, this helps them predict outcomes,
or create hypotheses about outcomes to guide them in further research,
if need be. For example, a business may consider testing the reactions of
its competitors should it raise its costs on various offers. Aside from
discovering if this yields a profitable change, it can discover how
companies in the same niche respond and if those responses drive more
sales, etc.
Key Independent Variables
Prices, Digital user experience such as new site features, Advertisements,
Marketing activity (social media announcements, retargeting, etc.),
Season, Inventory (new products or upgrades), Interactions with sales
agents, Key Dependent variables, Sales, Demand, feedback (whether
positive or negative), Site traffic, In-store visits, and Revenue.
An Example of Experimental Research for Business
Market researchers can apply experimental research to a wide breadth of
testing needs. Virtually anything that requires proof, confirmation, or is
clouded by uncertainty can put experimentation into practice. The
following is an example of how a business can use this research:
A product manager needs to convince the higher-ups in a denim company
to launch a new product line at a particular department store. The
objective of this launch is to increase sales, expand the company’s floor
presence and widen the offerings.
The manager must prove that this line is needed for the company to pitch
the idea to the department store. The product manager can then conduct
experimental research to provide a strong case for their theory, that a new
line can raise sales.
The product manager performs experimental research by executing a test
in a few stores, in which the new line of denim is sold. These stores are
varied in location to signify the target market sales before and after the
launch. The test runs for a month to determine if the hypothesis (the new
line resulting in increased attention and sales) can be proven.
8. This represents a field experiment. The product manager must heed the
sales and foot traffic of the new product line, paying attention to spikes in
revenue and overall sales to justify the new line.
Experimental Research Survey Examples
Survey research runs contrary to experimental research, unlike the other
main forms of research such as exploratory, descriptive, and correlational
research. This is because the nature of surveys is observational, while
experimental research, as its name signifies, relies on experimentations,
that is testing out changes and studying the reactions to the changes.
Despite the contrast of survey research to experimental research, they are
not completely at odds. In fact, surveys are a potent method to gain further
insight into an existing experiment or understand variables before
conducting an experiment in the first place. As such, businesses can
adopt a wide variety of surveys to complement their experimental
research. Here are some of the key forms of surveys that work in tandem
with experimentation:
The quantitative survey Discovers the aspects of statistical significance
within variables. Helpful in that causal research is quantitative in essence.
The retrospective survey. Delves into past events, occurrences, and
attitudes regarding the variables. Shows whether the variables changed
and how so. The prospective survey. Can find causative elements
between variables over a period. Useful for formulating hypotheses. The
customer experience survey. Helps businesses zero in on variables that
contribute to or result from certain kinds of customer experiences.
Measures various matters critical in a business or organization; surveys
employees. Deployed more frequently, so variables can always be
continually tracked.
The qualitative survey Helps answer what, why and how with open-
ended questions. Extracts key high-level information in depth. How
Experimental Research Differs from Correlational, Exploratory,
Descriptive and Causal Research Experimental research differs from
exploratory, descriptive, and correlational research in self-evident ways. It
is, however, often conflated with causal research. However, they too have
notable differences. Causal research involves finding the cause-and-
effect relationships between variables. Thus, it too employs
experimentation. However, this means that causal research is a form of
experimental research, not the other way around. Experimental research,
on the other hand, is fully science and experiment-based, as it chiefly
9. seeks to prove or disprove a hypothesis. While this largely involves
studying independent and dependent variables, as it does in causal
research, it is not solely based on these aspects. Instead, it can introduce
a new variable without knowing the dependent variable or experiment on
an entirely new idea (as in the example used in the previous selection).
Causal research investigates the comparison of variable relationships to
find a cause and effect, while experimental research states an expected
relationship between variables and is bent on testing a hypothesis. As far
as comparisons to correlational research go, while experimental research
also studies the relationships between variables, it functions far beyond
this by manipulating the variables and virtually all subjects involved in
experiments. On the contrary, correlational research does not apply any
alterations or conditioning to variables. Instead, it is a purely observational
research method. As such, it merely detects whether there is a correlation
between only 2 variables. In contrast, experimental research studies and
experiments with several at a time. Exploratory research is vastly different
from experimental research, as it forms the very foundation of a research
problem and establishes a hypothesis for further research. As such, it is
conducted as the very first kind of research around a new topic and does
not fixate on variables. Descriptive research, like exploratory research
and unlike experimental research, is conducted early in the full research
process, following exploratory research. Like exploratory research, it
seeks to paint a picture of a problem or phenomenon, as its zeros in an
already-established issue and delves further, in pursuit of all the details
and conditions surrounding it. Thus, unlike experimental research, it only
observes; it does not manipulate variables in any capacity or setting.
The Advantages and Disadvantages of Experimental Research
Experimental research offers several benefits for researchers and
businesses. However, as with all other research methods, it too carries a
few disadvantages that researchers should be aware of.
The Advantages
Researchers have a full level of control in an experiment.
It can be used in a wide variety of fields and verticals.
The results are specific and conclusive.
The results allow researchers to apply their findings to similar
phenomena or contexts.
It can determine the validity of a hypothesis or disprove one.
10. Researchers can manipulate variables and use them in as many
variations as they desire without tarnishing the validity of the
research.
It discovers the cause and effect among variables.
Researchers can further analyse relationships through testing.
It helps researchers understand a specific environment fully.
The studies can be replicated so that the researchers can repeat
their experiments to test other variables or confirm the results again.
The Disadvantages
It involves a lot of resources, time, and money, as such, it is not easy
to conduct.
It can form artificial environments when researchers unwittingly
over-manipulate variables as a means of duplicating real-world
instances.
It is vulnerable to flaws in the methodology, along with other
mistakes that can’t always be predicted.
Flawed experiments may require researchers to start their
experiments anew to avoid false calculations, measuring results
from artificial scenarios or other mistakes.
Some variables cannot be manipulated, and some forms of
research experiments are too impractical to conduct.
How to Conduct Experimental Research
Experimental research is often the final form of research conducted in the
research process and is considered conclusive research. The following
explains the general steps required to successfully complete experimental
research.
I. Identify your research subject, a question surrounding it and its
variables.
II. Form a specific research question.
III. Gather all available literature and other resources around the
subject.
IV. Conduct secondary research around the subject and primary
research via surveys.
V. If the topic involves a research process you have already begun, for
instance in exploratory, descriptive, correlational, or causal research
type, gather together the facts you already have and hand.
11. VI. Consider how they relate to your question and how they line up with
the secondary research you conducted.
VII. After your initial studies, form a hypothesis.
VIII. Design a controlled experiment.
IX. First, decide which variable(s) is dependent/ independent (if it
doesn’t involve experimenting).
X. Decide how far to vary the independent variable.
XI. In the experiment, manipulate the independent variable(s).
XII. Measure the dependent variable(s) while you study the independent
variable(s) alongside.
XIII. Make sure to control potential confounding variables.
XIV. Assign subjects to their designated experimental treatment groups.
XV. Keep the study size in mind; a larger study pool creates statistical
findings.
XVI. Assign your subjects to “treatment” groups randomly, with each to
receive a different level of “treatment.”
XVII. Use a control group, which receives no manipulation. This shows
you the test subjects as they appear/behave without any
experimental intervention.
XVIII. There are 2 types of groups for assigning your subjects:
XIX. A completely randomized design vs a randomized block design.
XX. Completely randomized design: every subject gets randomly
assigned to a treatment.
XXI. Randomized block design: aka stratified random design, subjects
get first grouped based on a shared characteristic, then assigned to
treatments within their groups at random.
XXII. An independent measures design vs a repeated measures design.
XXIII. Independent measure: subjects receive only one of the possible
levels of an experimental treatment.
XXIV. Repeated measures design: every subject gets each of the
experimental treatments consecutively, as their responses are
measured. It also refers to measuring the effect of an emerging
effect over time.
XXV. Continue experimenting on variables as needed, take
measurements, and take notes.
XXVI. Based on your experiment(s), put together a logical conclusion. It is
possible that it may need testing over time.
12. Conclusion
Although experimental research can be very complex, this research
method is the most conclusive. Using a scientific approach, it can help
you form tests on various business matters. While it is critical for
understanding your target markets and customers’ existing behaviours, it
can also be used to experiment on a wide variety of other matters. Before
launching a new product, or an updated one, for example, you can
conduct an experiment to understand the product in action. This helps you
avoid any glitches or undesirable qualities that will incur problems for your
customs and a bad reputation for your brand. Experimental research is
not for every business, yet if you decide to implement this form of
research, consider using surveys in tandem. An online survey platform
can help you establish and distribute your surveys to a wide network via
organic sampling to avoid biases. Although it is not a requirement, in
today’s age of excelling in customer experience, it is of the essence to
have as much data on your target market as possible. An online survey
tool makes this possible.
13. REFERENCE
1. Apel, R., Bushway, S. D., Paternoster, R., Brame, R., & Sweeten,
G . (2008). Using state child labor laws to identify the causal effect
of youth employment on deviant behavior and academic
achievement. Journal of Quantitative Criminology , 24, 337– 362.
2. Berk, R. A., & Rauma, D. (1983). Capitalizing on nonrandom
assignment to treatments: A regression-discontinuity evaluation of
a crime-control program. Journal of the American Statistical
Association , 78, 21– 27.
3. Dunning, T. (2012). Natural experiments in the social sciences: A
design-based approach. New York: Cambridge University
4. Farrington, D P., & Welsh, B. C. (2005). Randomized experiments
in criminology: What have we learned in the past two
decades? Journal of Experimental Criminology , 1, 9– 38.
5. Gottfredson, M. R., & Hirschi, T. (1990). A general theory of
crime. Stanford, CA: Stanford University Press.
6. Kirk, D. S. (2009). A natural experiment on residential change and
recidivism: Lessons from Hurricane Katrina. American Sociological
Review, 74, 484– 505.
7. Murnane, R. J., & Willett, J. B. (2011). Methods matter: Improving
causal inference in educational and social science research. New
York: Oxford.
8. Sampson, R. J., & Laub, J. H. (1993). Crime in the making:
Pathways and turning points through life. Cambridge, MA: Harvard
University Press.
9. Shadish, W. R., Cook, T. D., & Campbell, D.
T. (2002). Experimental and quasi-experimental designs for
generalized causal inference. Belmont, CA: Wadsworth Cengage
Learning.