Overview of Methodologies and Research Designs

Overview of Methodologies and Research Designs
The following will present an overview of research methodologies and designs used in the social sciences. The main objectives of this document outlined below.
Objectives
After a brief review, you will be able to discuss the:
• Characteristics of the three core methodologies.
• Basic designs available within the methodologies.
This basic and introductory documentation is no substitute for courses in research methods; we can only attempt a succinct and cursory introduction here. Our goal, however, is to provide enough information that you will be able to complete any relevant required coursework and get a start on:
• Writing the “Basic Design” section of your Research Plan, a.k.a the Scientific Merit Review Form.
Let’s get started.
The Core Methodologies
There are three methodologies common to doctoral research. They are:
• Qualitative methodology.
• Quantitative methodology.
• Mixed methodologies.
As you may remember, quantitative analysis is based on the positivist paradigm, whereas qualitative analysis is based on the interpretivist paradigm. Mixed methodologies, obviously, requires that the two paradigms be integrated, which is intellectually and philosophically quite challenging.
Before going on, please be aware that mixed-methods research, using both qualitative and quantitative methods and procedures, is complex and arduous, and can require special permission, additional coursework, and a mentor with expertise in mixed methods. Further, mixed-methods doctoral capstones usually take longer and therefore are more costly, both in time and money. Each of Capella’s doctoral schools has specific requirements for learners wishing to pursue mixed designs. You can find out more about the research guidelines of your school by visiting the PhD page and scrolling down to the School of Public Service Leadership heading.
What Is a Research Design?
If you say you are doing a quantitative study, you are saying very little. All that means, essentially, is that you will use quantified data—numbers. A research project, however, has to have a research design, which consists of a plan or blueprint that has a series of steps which must be carefully planned ahead of time. These steps flow, one into the next, in a necessary order. This order is the same whether you do a quantitative study, a qualitative study, or a mixed study. Of course, there are some exceptions—there are always exceptions in social science!—but the order is almost always the same.
The design describes, graphically, what you will do and how you will do it.
The “what you will do” are the steps. The “how you will do it” is described in methods and procedures.
But before we go into that, let’s outline what the main steps of any research design include.
The Main Elements or Steps in a Research Design
You have already encountered the first set of steps—the conceptualization of the study based on a literature review. I’ll review them quickly as a refresher. First, you describe a research topic. You then do a thorough literature review to support and justify the topic, on the basis of which you next craft a problem statement. Following this, you write a research question.
Now you are ready to choose a methodology and a research design. The design involves the following steps that will be the blueprint for answering the research question and solving the research problem, which, after all, is the core purpose of your study.
The first step in your design is to create a sampling plan. That is, a plan for obtaining your participants. We’ll deal with the details of the sampling plan later.
Next, you will devise a data collection plan. That is, a plan for gathering the information from those participants that you need to answer the research question. Following this, you will create a data analysis plan. This is where you actually come up with the answers to your research question. Last, you devise a plan for presenting the findings to other scholars.
These four elements are found in every research design. In studies with special needs, there may be additional elements, such as a risk mitigation plan if your participants are vulnerable people or if there are significant risks of harming the participants in some way.
So let’s look now at basic research designs found in quantitative studies.
Basic Designs in Quantitative Methodology
First, quantitative designs are the most common in the social sciences, so let’s start there. We’ll first divide quantitative methodology into three large families of quantitative designs. Different schools and writers use somewhat different terms, but at Capella University, we talk about three such families of quantitative designs:
• Experimental designs, which are the “gold standard” in quantitative studies, because they seek to test hypotheses about one or more variables, the independent variables, causing or influencing or predicting others, called the dependent variables. Experiments attempt to demonstrate that some variable or group of variables cause changes in another variable or group of variables. True experimental designs meet these criteria:
o Random assignment to groups. We’ll discuss randomization when we get to sampling later, but it means simply that the participants are randomly placed in the experimental group or in the control group. This is the best way to control against extraneous variables influencing the outcome.
o A control group, which is the group who do not receive the experimental condition. They might receive no treatment, as in a waiting list design, or they might receive a placebo treatment or a treatment-as-usual.
o An experimental group. This group actually experiences the experimental condition.
o At minimum, a pretest and a posttest to measure some dependent variable. Of course, some experimental designs are more complex than that.
• Next, we have the family of Quasi-experimental designs, which are similar to experimental designs in which there is some attempt to control for extraneous and confounding variables, but the control is less rigorous than in the experimental family. Quasi-experimental designs have two major types:
o Non-random control group designs, in which true random assignment cannot be done, but an effort is made to have a representative or matched group as a control, and
o Time series designs, a group of designs that include controls based on conducting numerous measures taken over a longer period of time to control for extraneous variability. We can’t go into more detail here, but your research methods texts provide a great deal of detail. Suffice it to say that time series designs are valuable when it is impossible to set up a non-random control group and the research question calls for a quasi-experimental design.
• Finally, we have the family of Non-experimental designs. Here there is no thought of a cause-effect relationship, as in the first two families. Non-experimental designs discover whether two or more variables are somehow related to one another. The most common kind of this design is the
o Correlational designs, in which the non-causal relationship between two or more variables are investigated. Generally, correlational designs are not approved for doctoral capstones, unless the number of variables is higher and some kind of predictive relationship is being sought by a multiple regression analysis of some sort.
o Difference designs, which analyze two or more groups’ scores on a dependent variable in order to determine whether there is a difference between the two groups on that variable. Again, as a doctoral capstone often ask for a higher degree of complexity than a simple two-group comparison on a single variable in projects using this design.
No doubt, if this is your first introduction to designs, this can be confusing. There is a handout showing the acceptable quantitative designs for your school. It shows common kinds of research questions that call for each of the various designs in these families.
Other Quantitative Designs
You should be aware that there are many other kinds of quantitative research designs that are used in social science. For example, meta-analysis is a powerful social science design. But these are the families we rely on in in general, because one goal of the doctoral capstone is to provide learners with solid experience using the basics of design. After graduation, if you go into professional research, you will be prepared to move to the next level of design sophistication.
There is one exception: A learner who wishes to do a more sophisticated design may receive approval if the mentor or a committee member has training and experience in the design, and the learner demonstrates additional coursework or training in the design.
Now, let’s move along to qualitative designs.
Basic Qualitative Designs
Just as in quantitative methodology, there are other qualitative research designs that are not usually used in our department. For the same reason, we use only these that have been approved by the department. And as with quantitative studies, if you wish to use a different qualitative design, you must have a mentor or committee member with training and experience using the design and you will have to demonstrate additional coursework or training in the design.
We approve six Qualitative Designs at Capella. They are:
• Qualitative case study.
• Ethnography.
• Generic qualitative inquiry.
• Grounded theory.
• Heuristics (also called heuristic phenomenology).
• Phenomenology.
Going into detail about each of these designs is beyond the scope of this document. However, you will be reviewing the handout “Design Diagrams,” which will show typical qualitative research questions and their related designs.