How do you go about conducting research? Saunders, et al. (2012) defined research design as a plan aimed at answering a specific research question. It concerns proper research data management and bringing together several components, strategies, and methods to collect data and analyze it. However, Bliesmer (1970) has said that designing research may fail to account for flexibility and the duration of the research.
In this article, we’ll look at the different types of research design and which particular field of study they are most suited to. We then outline an action plan to help enable readers to write a design plan of their own that is appropriate for a specific study, with the intention of making the design as comprehensive as possible.
A research design also called a research strategy, is a plan to answer a set of questions (McCombes, 2019). It is a framework that includes the methods and procedures to collect, analyze, and interpret data. In other words, the research design describes how the researcher will investigate the central problem of the research and is, thus part of the research proposal.
The design of this research influences the type of data to be gathered and, consequently, its results. Depending on the type, which we will explain below, research design also defines all other constituent parts of a study, such as variables, hypotheses, experiments, methodology, and statistical analysis (Creswell et al., 2018).
Many people confuse research design with research methodology, however. The difference is that while the former is an outline of how to approach the problem, the latter states how to implement the design. Both are crucial in building a thesis statement. And, if you need help with this part of your study, you can check out our guide on how to write a thesis statement for a research paper.
Excellent research design has one purpose: to make the data address the research problem as clearly, as accurately, and as unbiased as possible. Arriving at the results means successfully specifying the type of results to test a theory or evaluate or describe a phenomenon. Without doing this beforehand, interpreting data will appear weak and flimsy and will likely not address the problem the researcher has set out to answer.
In any research work, design is rudimentary since everything eventually emanates from the selected design, and since this selection is the most closely related to the scholar’s theories and research questions (Vogt et al., 2012). With the right choice, research design has fulfilled its purpose when the conclusion is seen to have a minimum bias. A research design that produces the least margin of error is one of its goals.
To do so, sound research design follows these main tenets:
de Vaus (2001) uses construction as an analogy for research design. Before constructing any building, knowing whether a builder needs a high-rise, a factory, a school, or any type of building is essential. Therefore, knowing the type and the characteristics of the research goes first naturally, before even beginning to pose the hypotheses and the methods used to collect data that, in turn, can support or abrogate the hypotheses. This approach is also very useful when writing your university dissertation.
The determination you make during the framing of the research design process will significantly dictate the value of the conclusions you can derive from your study outcomes (Bordens & Abbott, 2018). As such, ensure that your selected research design is highly appropriate because with the right choice comes the relevant results.
While the length and complexity of the research design vary, the research design itself consists of several parts. Note that the research problem will dictate the research design, including its type and its elements. These parts are:
The statistical methods to analyze such data can take on different forms and depend on the research’s central question. The chart illustrates which method is mostly used in the context of market research:
Source: NewMR, 2017
There are several ways to see that the study is designed well at a glance. Here are four main characteristics that make for good research design:
These four elements, including how the research is designed, influence how the research will be conducted and the methods used to acquire the results.
Research methods are used to answer different questions and are highly dependent upon the type of research design used to “frame” the entire study. Because they are so closely related, research methods are somewhat conflated with research design, but the subtle nuance is there. Yin (2014) has a succinct way of differentiating the two: design is logical, while method is logistical. In other words, the design is the plan, the method is how to realize that plan.
There are important factors at play when creating a research methodology. These include ethics, the validity of data, and reliability. Accounting for the time spent on collecting and analyzing data is also a prudent move. The type of research method will also factor into the time; for example, interviews or observation may yield rich data sets, but they take much more time than, say, a survey. Therefore, balancing these needs with the time and resources available and the advantages and disadvantages of each method will be paramount to designing a research method.
Alzheimer Europe (2009) outlines eight types of research methods, but other disciplines may demand more specialized ones. That said, the biggest contributor to the type of method will depend on the goals of the study. In the social sciences, for example, this may depend on the subject/s or the central problem, such as what makes people buy expensive designer clothing more than their more affordable counterparts.
The following methods, however, can be used in any field or body of knowledge:
Note that the research does not have to be pigeonholed into one particular type of method. Depending on the resources and the research design, the research team can combine several types of methods to find the data they need. In addition, the data generated from one method will be markedly different from that of another, both in quality and quantity.
The type of research design is one of the biggest contributors to the quality, relevance, and accuracy of a result. Therefore, before setting out to outline a proposal, it is always a good idea to distinguish the type of research by including it in the research design.
There are a few ways to approach a research design type, but literature has not always been clear-cut on these types (Abutabenjeh, 2018). In fact, existing publications have made it even impossible to distinguish between types, methods, and approaches, with some older references talking about fixed and flexible designs (Bouma, 1994).
To make designing research as simple as possible, we have broken down the types of research design into four major ones, as explained below.
In studies where the researcher is interested in describing a case, situation, or phenomenon, they are acting under a descriptive research design. As a theory-based design, it is interested in answering the how, what, when, and where questions, instead of the why. Descriptive research directs the researcher to understand the research problem before investigating why it even happens in the first place.
Descriptive design furnishes the researcher with an opportunity to gain insight into the problem itself. It also helps the research team to see the need for the research. If it is not as clear or as necessary, exploratory research (which, according to Blaikie (2000), is considered as the first phase of research) may be needed. Descriptive research attempts to build on the groundwork made by exploration, such as providing additional information, filling in gaps in knowledge, or expanding it. Unique to descriptive research is that it also aims to collect as much data and information as possible.
An example of descriptive research is market research. An investor, for example, may need to look at the market, such as its current state, its trends, and so on. Descriptive research can answer all these questions for the investor, which is why market research is an investment in itself, as evidenced by the following graphic.
Source: Global Market Research 2019
Using an experiment, the research attempts to establish a cause-and-effect relationship in a situation or phenomenon. It is a causal research design type where the researcher tries to observe the impact of a variable on a dependent one. In doing so, the researcher attempts to determine or predict what may occur based on experimental models (Anastas, 1999).
In the example above, experimental research entails the investor modifying a variable to look at how such a change affects other factors, such as price.
Experimental research is a practical route to take, as it allows the researcher to find exactly what is working and what is not, and account for these changes accordingly to solve the research problem. Experiments are often used in the social sciences and in the medical field by grouping people, such as by using a control group as an independent variable.
Like experimental research, correlational research aims to establish a relationship between two variables. The difference is that while experimental research tries to monitor changes between variables (causal), correlational research tries to look for associations and similarities between them (Sassower, 2017).
As a non-experimental technique, it instead relies on evaluating the relationship between these variables using statistical analysis. To calculate the amount of correlation between two variables, a statistical method called Pearson’s correlation coefficient is used (Mukaka, 2012), which is a value between -1 and +1. The more it leans toward a positive value, there is indeed a relationship between the two. A negative value denotes the variables are related but indirectly proportional, and zero denotes no relationship.
As evidenced by the name, explanatory research aims to explain the researcher’s findings and ideas to expand the theory. Using this research design, the researchers explore the limits and boundaries of a subject in order to present the reader with the results that answer the what, how, and why of the research’s central thesis. When conducting the research, the researcher should leave all biases behind and adapt to new data and/or findings.
Researchers and students conduct explanatory research to find the underlying problem or a new angle to a problem. These may not always be readily apparent when initially proposing the research or it was not studied in-depth before (GradesFixer, 2019).
Note that explanatory research does not seek to provide conclusive answers, but to give an avenue to researchers to plumb the depths of the subject.
Any researcher would need an understanding of the research design types to see which is more appropriate for the study or which one brings to fruition the most accurate results. To do this, there are three broad ways to approach the design, as we will discuss below.
The first approach is to design the research using a quantitative perspective. This approach best suits a research goal where actionable insight is tied to a statistical conclusion. As the name implies, the quantitative approach frames numbers as a representation of data. Because numbers are objective, a quantitative approach is necessary, for example, in making data-driven business decisions where margins of profits turn on the most minute of details and/or figures.
Often called a “top-down” approach (Burney, 2008), it involves taking away the parts from the general to the specific. This way, the researchers arrive at a conclusion based on the premises or the available facts. Because quantitative data also tend to be voluminous, statistical software and other services are used to analyze them, the most popular of which are illustrated below.
Source: NewMR, 2017
Objectivity is highly prized in a quantitative approach to research. As a result, researchers go to great lengths so that the results of their research are untarnished by their own presence, behavior, or expectations. One such way to do so is by self-examination such that their methods or conclusions are free of unwarranted biases or presumptions. A quantitative approach to research must mean that the design itself must account for and/or control external variables; they can never be eliminated completely, and as such must be acknowledged in the interpretation of the findings.
In order to approach a design with quantitative intent, researchers often start with one or more hypotheses and the relationships between the variables they want to investigate. The design must also factor in stricter forms of methodology and tools used to measure and validate the collected data, a clear plan of action, a statistical procedure to analyze data, and a valid way to present these results.
A qualitative approach, on the other hand, to research sets out to determine a relationship between collected data and observations. As it is about recording, analyzing, and discovering the web of interconnectedness that underpins related subjects, it generates a plethora of raw data, whether obtained through statistical means or otherwise. This nature of a qualitative approach thus lends itself well to exploratory research (Blaikie, 2009).
Unlike quantitative approaches, a qualitative approach is its opposite. It uses an inductive way of approaching the conclusion of the study. Also called a “bottom-up” approach, it infers meaning or looks for patterns on the basis of the data that they have collected.
Qualitative research is employed extensively in the social sciences. It is concerned with observing and uncovering the social constructs that human societies are framed in and looks at the significance of the human experience in the lens of beliefs, behaviors, and emotions. A qualitative approach is, thus interested in gaining an understanding of what works as the participant sees and feels it (that is, subjectively).
While qualitative research also uses some form of quantitative analysis, the way it collects data allows for greater freedom. Unlike quantitative methods that gather and collate data in computer-readable forms to be crunched at a later time, qualitative approaches record data in textual format from observation and interaction with the subjects. In addition, the methods used vary wildly, with open-ended, exploratory, and wide-ranging processes and little, if any, assumptions on the part of the researchers so as to make the data pristine and as accurate as possible.
Ultimately, quantitative and qualitative methods should not be considered as strict, divergent dichotomies, opposites, or categories. Rather, they should be seen as representing distinctive ends on a continuum or large system (Creswell, 2015). For instance, a research work tends to be more quantitative than qualitative or the other way around (Creswell & Creswell, 2018).
Aristotle himself probably initiated the first pragmatic approach to research (Teddlie, 2008), though this is more of a proto-mixed method. A true pragmatic researcher, instead, uses any or both approaches to research design as fits the scope of the study and the questions it seeks to answer.
True to its name, the pragmatic approach cares nothing about which approach is better or in the philosophical ramifications of choosing one over the other. Instead, it uses the method that appears to best suit the task at hand. As a result, pragmatists use whatever tools, reasoning, and techniques that are appropriate to the situation without worrying over what kind of approach they are doing. Central to this idea is that they recognize that each approach has its attendant pros and cons, but they can be complementary as well.
Data that pragmatists collect this way are measured and analyzed using the appropriate manner (for example, a qualitative literature review is examined qualitatively, while a statistical survey is done quantitatively). The advantage, however, is that data can be converted between these two measures (especially qualitative to quantitative).
As this article has explained, a research design is independent of any method or procedure for collecting data. By definition, any type of research can do so, and researchers can perform any type of approach as appropriate to the situation or the problem of the research. What research design aims to do is to create a direction or a blueprint of the research by recommending a framework of the inquiry.
The design of the research thus enables researchers to identify the type of data collected and the evidence they need to answer the question. Simply collecting evidence to support a postulated hypothesis would not do; it is the task of the researcher, by designing a framework, to allow the study to find alternative, even conflicting explanations and which one makes more sense or can be most validated.
Scholars and researchers use a variety of research methods to collect and analyze information. To know more about these methods, read our article on primary and secondary research methods.